Track-Guided Hierarchical Reinforcement Learning for Autonomous Vehicle Drifting with Minimum-Lap-Time Planning
基于轨迹引导的层次强化学习在自主车辆漂移中的最小圈速规划
Zhao, Sheng, Zhao, Bolin, Wu, Xiaodong, Lv, Chen
Abstract
In Formula 1, drivers optimize racing lines within tire grip limits to minimize lap times; however, in rally racing, drivers intentionally break traction to drift on loose surfaces. This maneuver rapidly aligns the vehicle for corner exits, ultimately reducing lap time. Autonomously executing such maneuvers formulates a complex dual-objective control problem: stabilizing highly nonlinear drift dynamics while strictly minimizing lap time. Addressing this challenge motivates the development of advanced Minimum-Lap-Time (MLT) drift control architectures. This paper proposes a planning-control framework specifically designed for MLT drifting scenario. First, we formulate an optimal control problem to generate a MLT drift planning trajectory, which is used as prior data to train a deep reinforcement learning drift controller. Given that drifting involves extremely large sideslip angles and is therefore challenging to learn directly, a Track-guided Reinforcement Learning (TgRL) drift control method is proposed to enable progressive training in a step-by-step manner, from drift control policy, to drift corner policy, and finally to a comprehensive drift race policy. The reward function incorporates both an instant reward term and an end reward term derived from the Minimum-Lap-Time objective. Simulation results demonstrate that the proposed framework enables the agent to learn a drift racing policy that not only ensures vehicle motion control performance but also effectively reduces lap time.
Navigating doorways is a fundamental capability for mobile manipulators operating in human environments, requiring coordinated motion between the mobile base and manipulator arm. This paper presents a motion planning framework that generates dynamically feasible and collision-free trajectories for autonomously opening and traversing both push and pull doors. The proposed method formulates the robot and door as a coupled dynamical system within a nonlinear Model Predictive Control (MPC) optimization framework. Manipulation feasibility is enforced through a penalty-based constraint, avoiding explicit arm kinematic modeling in the planner. Simulations and a hardware experiment demonstrate that the approach successfully plans feasible trajectories for door traversal.
Developing Combined Manipulation and Locomotion Skills with Interaction Representation and Skill Composition
结合交互表示和技能组合开发操控与运动技能
Meng, Fanxing, Xiao, Jing
Abstract
This paper addresses how to enable a humanoid robot to learn motion policies based on developmental principles and combine policies to create more sophisticated and useful behaviors. Specifically, we present an approach to (1) learning a whole-body reaching and grasping policy and (2) combining it and a standing-up and walking policy to compose a more complex policy of manipulation and locomotion: grasping, standing up, and walking. In (1), our method draws inspiration from harmonic analysis and adopts cubic harmonics as weights to represent the hand-object spatial relationship via spatial convolution. Utilizing an intra-episode finger joint decoupling curriculum based on developmental principles, a robot can autonomously learn a generalizable grasping policy without relying on external datasets or pretrained models. In (2), our method combines the grasping policy with a separately learned getting-up policy by providing both policies with their respective observation vectors and using hand-object interaction scores to determine when each policy should control which robot joints. Our results show a 93% zero-shot success rate for grasping unseen objects and a 96-100% success rate for standing up while holding the object. Our work also demonstrates that combining different policies is only effective if each policy learning happens on the same whole humanoid body even if a policy (such as for locomotion) does not seem to need all the body parts (such as fingers).
Contact planning in robotics requires models that are both computationally efficient and physically accurate. Sphere-based robot representations satisfy the first requirement by enabling fast collision checking and differentiable geometry, but sacrifice physical accuracy by relying on point contact which cannot capture contact patch area, pressure distributions, rotational stiffness, or frictional moments. We introduce Compliant Sphere Lattice Contact (CSLC), a distributed contact model that operates natively on sphere representations by modeling the robot interface as a compliant lattice of surface spheres connected through anchor and lateral springs. When pressed against an object, the lattice deforms to produce a spatially distributed contact patch that improves the physical accuracy of sphere-based contact. We validate CSLC across two independent solvers and show preliminary results demonstrating contact patch formation and improved grasp stability.
Localization in Spatiotemporal Fields via Environmental PDEs
通过环境偏微分方程进行时空场中的定位
Fuentes, Jose, Newaz, Abdullah Al Redwan, Cavalcanti, Ana, Bobadilla, Leonardo
Abstract
This paper proposes a localization framework that uses spatiotemporal fields governed by partial differential equations (PDEs) as localization signatures. Two PDE classes are considered: the shallow water equations, which describe free-surface flows in coastal and riverine environments, and the advection-diffusion equation, which models the transport and mixing of scalar quantities such as temperature, salinity, and dissolved oxygen. A numerical PDE solver provides predicted fields over the domain, and multiple field channels are fused as multimodal measurements to improve localization accuracy. We formulate the problem within a Rao-Blackwellized particle filter (RBPF) that partitions the vehicle state into a nonlinear component sampled by particles and a linear sensor bias component tracked analytically via per-particle Kalman filters. This factorization reduces the required number of particles compared to a standard particle filter while accounting for realistic sensor drift. Simulation studies on both PDE scenarios show that the RBPF consistently outperforms a standard particle filter in terms of final position error and Root Mean Square Error (RMSE) across varying particle counts. Field experiments with an autonomous surface vehicle measuring salinity, temperature, and dissolved oxygen validate that PDE-governed environmental fields provide sufficient spatial variability for practical localization. Related experimental videos are available at https://localization-environmental-pdes.github.io/.
This paper presents an applied case study on hybrid visual attention estimation for human-robot interaction using an expressive robotic head based on the InMoov ecosystem. The proposed pipeline combines a fast geometric perception layer with an independent semantic perception layer based on a vision-language model. The geometric layer provides high-frequency face and head-pose information for temporal regulation, while the semantic layer receives only raw egocentric camera frames and produces contextual attention labels related to attention toward the robot, phone use, or attention elsewhere. These signals are integrated through a finite state machine that regulates adaptive interaction behavior, including activation, waiting, interaction resumption, and return to rest. The system was evaluated with 10 participants across 40 trials covering baseline and adaptive interaction conditions. Results show reliable interaction start across all trials, consistent pause behavior in the adaptive distraction condition, and non-redundant semantic information between the geometric and semantic outputs.
Towards General Language-Conditioned Latent Safety Filters
面向通用语言条件的潜在安全过滤器
Tabbara, Ihab, Yang, Yuxuan, Sibai, Hussein
Abstract
Robot policies are becoming increasingly general, with vision-language-action (VLA) models enabling a single policy to execute diverse tasks specified in natural language. Safe deployment, however, requires adapting not only to new tasks but also to varying safety requirements across users, environments, and applications. Existing safety filters remain largely constraint-specific and thus must be redesigned or relearned when safety requirements change. In this paper, we investigate language-conditioned safety filtering, in which a Hamilton-Jacobi safety actor and critic are conditioned on language-specified constraints. We evaluate this formulation across pick-and-place, table-wiping, and block-stacking tasks in the vision-based setting, examining its ability to enforce language-specified constraints and transfer to unseen constraint instances within the evaluated constraint families. Our experiments provide evidence that language-conditioned safety filters reduce constraint violations and exhibit partial transfer to unseen constraint instances.
Belief-Space Perception Routing under Coupled Sensor Faults and Compute Contention
在耦合传感器故障和计算竞争下的信念空间感知路由
Roy, Sparsh, Aggarwal, Vihan, Yin, Davin
Abstract
A robot that has to see and react on a fixed clock runs into two problems at once. Its cameras degrade in rain, mud, fog, and darkness. And the single onboard processor it runs on is shared with planning and control, so the compute left over for perception moves around from second to second. Most systems model the two separately. We present a perception router that tracks probabilistic estimates of sensor-fault state and compute- contention state, couples them with a noisy-OR term, and uses the coupled estimate to pick one of four detector configurations (YOLO11x/n at 1280 or 640 px) so that the frame finishes before its deadline. Where the two stressors co-occur, the coupled policy cuts the deadline-miss rate by 1.1 to 9.4 percentage points against a policy that treats them independently. The interval excludes zero in five of six conditions, the pooled effect over 10 sequences and 6 conditions has sign-test p = 0.001, and every uncoupled control and the fault-free trajectory sit at exactly 0.0 pp. Routing costs tens of microseconds per frame. We then asked whether the coupling the method exploits arises on its own. Across eight real RADIATE adverse-weather sequences and three workload proxies independent of the fault signal, after Benjamini-Hochberg correction and a replication run, none of 24 tests found it. We report that null and scope the routing result as a proof of mechanism. Whether such coupling occurs in the field is still open, and the released evaluation pipeline lets a deployment settle it on its own traces.
Deploying robot foundation models at scale is the next step towards realizing the potential of general-purpose robots. However, Vision-Language-Action (VLA) and other foundation models are computationally demanding, and on-device compute is constrained by power and space. In this paper, we introduce the problem of serving a robot policy to multiple robots from a remote GPU and formulate it as a scheduling problem. We build Armory, a serving system validated on fleets of both simulated and real robots. Our experiments show that naive scheduling heuristics perform well when all robots are the same, but fall short when robots consume action chunks at different rates, uncovering a mismatch between conventional batching methods and the closed-loop requirements of robot policy execution. To address this, we propose a scheduling algorithm that accounts for this heterogeneity and improves overall system throughput by up to $18\%$ in real-world experiments. Additional details are available at https://gatech-rl2.github.io/actionchunkscheduling.
A Robotic System for Automated Manufacturing of Dielectric Elastomer Actuators
用于介电弹性体驱动器自动化制造的机器人系统
Remenar, Van, Gogoj, Anatol, Li, Ang, Verardi, Tomas, Mandoiu, Matei, Santiago, Victor Jimenez, Duduta, Mihai
Abstract
This letter presents an automated robotic manufacturing system for soft capacitors which operate as actuators and sensors. Emphasis is placed on the two processes that most directly govern device quality, dielectric layer formation by spin coating and carbon nanotube (CNT) electrode application by stamping. Twenty multilayer DEAs, each comprising 12 dielectric layers with a mean thickness of 55.37 +- 2.04 um and 11 alternating CNT electrodes, were fabricated reducing total process time by 14.2% and removing the operator from 56.1% of it.
Token skipping is a widely used training-free way to accelerate vision--language--action (VLA) models by bypassing computation for most visual tokens at each control step according to a gate. When the next gate is harvested from the previous accelerated forward, however, the tokens skipped at one step are also the ones least visible to the next gate, and the damage can compound across control steps until the task fails. We study the two mechanisms this class is built on, reuse and deletion, crossing each against where its gate signal comes from on identical episodes. At a skip ratio of 0.9 on LIBERO-Object, both collapse when the gate comes from the model's own accelerated forwards, to 0.68 under reuse and to 0.31 under deletion against a dense 1.00, and the collapse is invisible to the action-level detectors we evaluate. What separates collapse from dense-level operation is not the mechanism but whether the gate is clean, computed by a forward that skipped nothing. We therefore propose actuation-slack refresh, one dense pass run while the robot executes its current action chunk, off the critical path, that hands the next step a clean gate and a fresh KV base. Since the measured detectors do not reliably reveal the failure, the refresh is unconditional rather than triggered. Both mechanisms then recover to 0.98, keeping the speed of skipping and the information of a dense pass. We then integrate the refresh into state-of-the-art caching and pruning methods across two VLA policies, 4 LIBERO suites, and 4 SIMPLER tasks, where it repairs every collapse caused by using a self-harvested gate. Serve latency drops 18--22\% below dense, measured both in simulation and on a physical robot. Where the gate signal comes from, not how tokens are skipped, decides closed-loop reliability for accelerated VLAs.
Aerial robots generate structured aeroacoustic fields during flight, yet these signals have been underexplored as a source of onboard relative perception, particularly under the strong ego-acoustic interference generated during simultaneous flight in various outdoor conditions. We introduce embodied passive aeroacoustic perception, a sensing paradigm in which an aerial robot infers actionable relative-state information from the naturally generated sound of flight while operating within its own evolving aeroacoustic field. We present SonicFly, a passive aeroacoustic perception framework that enables one unmanned aerial vehicle to estimate and follow another using only the leader's intrinsic flight sound, without active acoustic signaling, inter-robot communication, GPS sharing, or external sensing infrastructure. The system uses a lightweight four-microphone array, rotorcraft-informed acoustic representations, a neural bearing-range estimator, and confidence-gated filtering for closed-loop flight. Through acoustic characterization, onboard localization, and outdoor pursuit experiments, we show that multirotor aeroacoustic signals contain sufficient information to support relative perception despite strong ego-acoustic interference, environmental variability, and changing flight geometry. During acoustic-only pursuit, SonicFly achieved a mean distance-maintenance error of 1.34 m across diverse outdoor trajectories and operating conditions. Analysis of the acoustic channel reveals design principles governing embodied passive aeroacoustic perception, including the roles of harmonic structure, spectral separability, and spatial acoustic cues in determining observability. Our results establish the feasibility of embodied passive aeroacoustic perception for aerial robots and suggest that naturally generated behavioral signals can serve as information for robotic perception and coordination.
Based on vision and prior experience, humans can make rough physical predictions and adjust their manipulation strategies. This paper aims to endow robots with a similar ability. To collect paired data of vision and forces, we use a rigid-body simulator commonly adopted in robotics. However, unlike simulators that output noisy point forces, humans are able to make consistent predictions even in unfamiliar situations. Based on this observation, we hypothesize that predicting smooth force distributions rather than raw point forces can improve both force prediction itself and downstream task performance. To validate this hypothesis, we construct a model that predicts three-dimensional force distributions from a single RGB image of piled daily objects. The target distribution is generated by applying statistical smoothing to point forces obtained from the simulator. Moreover, by incorporating object geometry into the smoothing process, we aim to account for variations in contact states and achieve more consistent vision-based predictions. We conduct extensive evaluations in both simulation and real environments. Results show that our approach improves prediction accuracy, enhances downstream task performance through smoothing, and further benefits from geometry-guided smoothing. Remarkably, the trained model generalizes effectively to real-world scenes despite being trained solely in simulation.
From Digital to Physical Reservoir Computing: Co-Optimizing Soft Robotic Reservoirs via Dynamics Matching
从数字到物理水库计算：通过动态匹配共同优化软机器人水库
Visentin, Nicola, Stölzle, Maximilian, Montero, Mariano Ramírez, Braghin, Francesco, Rus, Daniela, Della Santina, Cosimo
Abstract
Soft robotic substrates are promising for Physical Reservoir Computing (PRC) because their compliant nonlinear dynamics can provide temporal memory, high-dimensional state transformations, and efficient inference. However, physical reservoirs are often adopted as-is rather than pretrained or co-optimized, potentially limiting soft robotic PRC performance relative to digital reservoirs. We investigate whether a physical reservoir can instead be pretrained against high-performing digital reference dynamics. Our formulation jointly optimizes physical parameters, a diffeomorphic physical-reference state map, and feedforward-feedback control using a differentiable physical model and an acceleration-level equation-error objective that avoids temporal integration. As a proof of concept, we instantiate the formulation with simulated soft robots, a Random Oscillators Network (RON) reference, and parallel multi-start gradient descent. We evaluate the optimized reservoirs on classification (sMNIST and ADIAC) and forecasting (Mackey-Glass and Lorenz96) tasks across four reservoir dimensions. Compared with unoptimized soft robot reservoirs, the optimized reservoirs achieve a mean relative improvement of 33.7% across all tasks and datasets, while remaining close to the digital reference. These results demonstrate the feasibility of dynamics-level co-optimization for the simulated soft robotic reservoirs considered here.
Unified humanoid policies handle agile whole-body motion, yet stumble on a simple demand: staying balanced on one leg. On our single-leg-balance benchmark, eight released state-of-the-art general policies hold a clean single-leg stance on 0 of 90 test motions; they stay up only by stepping or hopping, recovering from imbalance rather than preventing it. Prevention needs the capture point (xCoM), the center of mass (CoM) extrapolated by its velocity, which has never driven a hardware policy because it requires a base linear velocity no on-board sensor provides; expressed relative to the support foot, that velocity cancels exactly, leaving an observation reconstructible from encoders and IMU alone. We put this first deployable dynamic-CoM observation directly into the actor that runs on hardware, and pair it with a reward library translated term by term from human postural control, under one principle: prevention over repair. Trained by asymmetric FastSAC with a privileged critic and no distillation, the resulting policy, FDDC (First Deployable Dynamic-CoM), holds clean single-leg balance on 86 of 90 held-out motions across nine stratified pose classes and transfers to a real Unitree G1; in ablation, the dynamic-CoM observation is the single largest driver: removing it alone costs 40 points of clean single-leg balance. We release the full stack with the first method-agnostic, reproducible sim2sim benchmark for humanoid single-leg balance, scoring each policy in a simulator distinct from its training one, a step toward turning balance from a per-task trick into a capability the field can measure.
Chinese Translation
统一的人形政策能够处理灵活的全身运动，但在一个简单的要求上却表现不佳：保持单腿平衡。在我们的单腿平衡基准测试中，八个发布的最先进通用政策在90个测试动作中没有一个能够保持干净的单腿站立；它们仅通过迈步或跳跃来维持平衡，恢复失衡而非预防失衡。预防需要捕获点（capture point，xCoM），即通过其速度推算出的质心（center of mass，CoM），而这一点从未驱动过硬件政策，因为它需要一个基础线速度，而没有任何板载传感器提供；相对于支撑脚，该速度恰好抵消，留下的观察数据仅能通过编码器和惯性测量单元（IMU）重建。我们将这一首个可部署的动态重心观察直接放入运行在硬件上的执行器中，并与一个逐项翻译自人类姿态控制的奖励库配对，遵循一个原则：预防优于修复。通过不对称的快速软演员-评论家（FastSAC）训练，使用特权评论家且不进行蒸馏，得到的政策FDDC（首个可部署的动态重心）在九个分层姿态类别中对90个保留动作中的86个保持了干净的单腿平衡，并成功转移到真实的Unitree G1；在消融实验中，动态重心观察是最大的驱动因素：单独移除它会导致40分的干净单腿平衡损失。我们发布了完整的技术栈，并提供了首个与方法无关、可重复的模拟到模拟（sim2sim）基准，用于人形单腿平衡，在与其训练环境不同的模拟器中对每个政策进行评分，朝着将平衡从每个任务的技巧转变为该领域可测量的能力迈出了一步。
SSTG-Nav: Metric-Grounded Spatial-Semantic Topological Graphs for Reusable Object Navigation
SSTG-Nav：用于可重用对象导航的度量基础空间-语义拓扑图
Peng, Daojie, Wang, Bingtao, Ma, Jun
Abstract
Service robots operating for months in the same homes, offices, and facilities should become more reliable with experience instead of searching familiar space from scratch for every request. Yet ObjectNav is predominantly formulated as one-shot exploration, leaving a central deployment challenge unresolved: recognizing an object does not identify a reachable place to stop, and one confident map error can terminate the task. We introduce SSTG-Nav, a reusable metric-semantic memory that turns a one-time survey into actionable object goals, consolidates evidence across viewpoints, and retains spatially distinct recovery standoffs. On 1,000 HM3D-v2 episodes across 36 scenes, our goal-independent topology achieves a 99.4% geometric success ceiling. Holding semantic responses fixed, metric grounding raises SR/SPL from 0.835/0.560 to 0.920/0.603, and source-aware fusion reaches 0.926/0.586. Fusion-aware Top-3 recovery raises Success@1/2/3 to 0.928/0.965/0.975 and reaches 0.601 SPL@3. Model, field-of-view, density, and corruption controls identify where these gains originate, and a ROS2/Nav2 realization demonstrates the complete reusable query-to-execution pipeline. Together, the results establish pre-exploration as a powerful practical regime for dependable, repeated semantic navigation.
Disentangling Visuo-Tactile Foresight: Oracle-Guided Interface Discovery for World Action Models
解开视觉-触觉前瞻的奥秘：面向世界动作模型的Oracle引导接口发现
Yao, Zihang, Ding, Chaoyue, Yu, Yingying
Abstract
Contact-rich manipulation remains challenging because successful control depends on physical interaction cues that are often weakly observable from vision alone. Recent tactile world action models jointly model future visual observations and tactile signals to guide action generation, but how such futures should be structured for effective use by the action expert remains underexplored. Directly studying this question with learned world action models is difficult because end-to-end behavior entangles physically invalid visual futures, unreliable predictions, inaccurate or cross-modally inconsistent tactile forecasts, and an unreadable future-to-action interface. To make this interface independently studyable, we introduce Oracle Visuo-Tactile Foresight (OVTF), a controlled framework that supplies paired RGB and tactile futures from successful trajectories verified in simulation. By fixing the future provider, OVTF isolates the interface and asks a cleaner question: if the future is successful and physically executable, what representation allows the action expert to absorb its benefit? Within OVTF, we propose Asymmetric Phase-Local Future Memory (AFM), in which visual memory reads future vision, each tactile memory jointly attends to its own tactile stream and phase-aligned future vision, and cross-tactile access is blocked. We compare AFM with Modality-Isolated Future Memory (IFM), which removes visual-to-tactile access and processes each future modality independently. Across seven tasks on the UniVTAC simulation benchmark, AFM achieves 32.0% average success, compared with 23.7% for IFM and 14.9% for UniVTAC-ACT. This controlled comparison shows that selective phase-aligned visual-tactile routing provides a more actionable future-to-action bridge than complete modality isolation.
Dexterous object rotation is a sequential contact problem: each support, release, and re-contact decision must both produce the desired object motion, and prepare the hand configuration for continued rotation. Existing reinforcement learning methods discover such movement patterns through trial and error on specific robotic hand embodiments, without explicitly accounting for how each contact transition affects the hand's ability to sustain object rotation in subsequent steps. We introduce DexMani, a framework that transfers human demonstrations as contact-conditioned manipulability evolution. This prior captures how successful human contact transitions reshape the object-rotation directions available to the hand. DexMani then learns this manipulability evolution and uses it to guide downstream reinforcement learning, enabling rotation skills to be acquired across robot embodiments with distinct kinematics and active-contact configurations. Across the Shadow Hand, Allegro Hand, and XHand, DexMani achieves the highest success rates in every evaluated setting for both seen and unseen objects. DexMani reaches an average success rate of 57.5% on LEAP Hand, outperforming other baselines and producing smoother rotatory motions. Project site: https://dexmani.github.io
Latency-Tolerant Cloud-Edge Collaborative Vision-Language-Action Models via Emergent Representational Specialization
延迟容忍的云边协作视觉-语言-动作模型通过新兴表征专业化
Peng, Daojie, Ma, Fulong, Wang, Bingtao, Wang, Sheng, Ma, Jun
Abstract
Deploying billion-parameter Vision-Language-Action (VLA) policies on mobile robots creates a systems conflict: semantic reasoning benefits from cloud GPUs, whereas closed-loop control must respond locally despite network delay and jitter. Existing hierarchical and asynchronous policies improve throughput, but their slow-path representations can still arrive stale or require explicit scheduling and delay cues. We introduce CloudEdgeVLA, a cloud-edge policy that treats temporal misalignment as a representation-learning problem. A cloud VLA encodes delayed observations into slowly varying task features, while a lightweight edge head combines the latest available cloud feature with current local vision. During training, current and randomly delayed frames are paired with the same current action target in fresh and stale paths. This objective encourages the cloud representation to preserve task-level information while the edge path supplies state-sensitive corrections. Across four LIBERO suites, CloudEdgeVLA retains 63.8--78.0% success with a 40-step uniform-delay window, whereas VLASH reaches at most 6.4% and the evaluated single-path baselines at most 3.0%. By removing blocking synchronization from the control loop, the design offers a practical route to scalable VLA deployment in which cloud models can grow while edge computation remains lightweight and responsive.
Grasp Execution Without a Planner: Configuration-Space Grasp Distance Fields with Certified Safety & Guaranteed Quality
无需规划器的抓取执行：具有认证安全性和保证质量的配置空间抓取距离场
Enwerem, Clinton, Baras, John S., Belta, Calin
Abstract
Standard multifingered grasp execution architectures plan a collision-free trajectory to a selected grasp pose and track it with a feedback law. Execution-time object pose uncertainty or perturbations may invalidate the planned trajectory, forcing a costly replanning step. We present Grasp Distance Fields (GDFs), a smooth softmin distance field over grasp configurations in the arm-hand configuration space. Our controller executes a grasp by following this field's negative gradient with a stationary feedback law, eliminating the need for a planner, stored trajectory, or grasp selection. For safety, we filter this command through a CBF-CLF quadratic program (QP) constraining self-collision, workspace, object, and obstacle clearance, reporting impeded progress as explicit slack. We prove that the softmin tracks the true set distance within \(\log N/\rho\) for \(N\) candidates and smoothing parameter \(\rho\), and that the filtered loop renders the safe set forward invariant. Since no smooth field captures hand-object contact switching, we switch modes with hysteresis at a pregrasp configuration, trading object collision avoidance for contact admission. A wrench-quality barrier then keeps the realized grasp's force-closure margin within a prescribed tolerance of its value at hold onset. We evaluate in kinematic simulation on a fixed-base arm and a Unitree G1 humanoid, both fitted with the same underactuated hand. Our controller navigates cluttered and dynamic scenes to safely reach for, grasp, and lift 46 of 50 objects spanning primitive, household, and adversarial classes. Across those lifts, the executed grasps retain a median 94\% of their synthesized quality margin, at 0.09 ms QP solve time per 20 ms control step. Per-step softmax weights confirm that our controller executes the nearest candidate, obviating a separate selection step. Project page: www.clintonenwerem.com/gdf.
From Failures to Supervision: DynamicEnvPlan for Robust Long-Horizon Embodied Planning
从失败到监督：动态环境规划（DynamicEnvPlan）用于稳健的长时间段体现规划
Yuan, Hao, Wang, Yuxin, Ji, Lei, Yu, Zhiwei
Abstract
Physical-world interaction is inherently dynamic, as environments can evolve during execution, requiring agents to adapt their plans under non-stationary conditions. We study this challenge through long-horizon embodied planning under environment deviations and execution uncertainty. Existing embodied-task benchmarks can expose such failures, but these failures are usually treated as evaluation outcomes instead of learnable signals for training agents to recover. In this work, we introduce DynamicEnvPlan, a closed-loop framework for high-level planning in dynamic environments. It extends embodied task execution with humanoid agents, high-level primitive skills, structured semantic memory, and controllable perturbations. Our data synthesis design consists of planning, perturbation, and guarded correction modules that turn dynamic execution states into recovery-oriented traces. The resulting traces are used for staged supervised fine-tuning, enabling the planner to learn from both nominal execution and perturbed recovery trajectories. Using 104 task-scene combinations spanning i.i.d., compositional generalization, and out-of-distribution settings for fine-tuning and evaluation, DynamicEnvPlan boosts success rate from 33.3% for the base planner to 76.2%, while improving across all seven evaluation metrics critical to physical-world interaction, including safety and affordance compliance.
Learning-Based Motion Planning for Dynamic Environments: From Foundational Algorithms to Emerging Paradigms
基于学习的动态环境运动规划：从基础算法到新兴范式
Shen, Zongyuan, Gupta, Shalabh, Zhao, Shancheng, Zhou, Dehua, Wang, Gao, Cheng, Rui, Ou, Yaming, Ren, Zhongqiang, Zhai, Yikui, Chen, C. L. Philip
Abstract
Motion planning in dynamic environments is a fundamental problem in robotics, aiming to generate safe and efficient paths, trajectories, or control actions in the presence of moving obstacles, uncertain predictions, and multi-agent interactions. It has broad applications in autonomous driving, service robotics, warehouse logistics, human-robot collaboration, crowd navigation, and multi-robot systems. This survey reviews representative works published primarily between 2015 and 2025, with a particular focus on how recent learning-based advances extend, complement, or interact with classical planning foundations. We first revisit classical planning methods as algorithmic foundations and reference frameworks for learning-based extensions. We then propose a role-of-learning taxonomy that categorizes existing methods according to how learning participates in the planning pipeline, including direct policy learning, learning-augmented classical planning, hybrid planning, and training enhancement methods. For each category, we summarize the main problem settings, representative algorithms, key ideas, integration mechanisms, strengths, and limitations. We further analyze how observation representations, prediction uncertainty, interaction modeling, planner integration, safety constraints, and training strategies shape learning-based motion planning in dynamic environments. Finally, we discuss open challenges and future directions, including sim-to-real gap, safe and certifiable planning, dense crowd navigation, perception-planning coupling, and embodied AI.
We present FlowPilot, a compact world-action model for real-time onboard UAV navigation from depth. Unlike map-then-optimize pipelines that require local reconstruction or end-to-end policies that lack explicit scene prediction, FlowPilot jointly denoises future depth observations and executable trajectories with flow matching. A dual-stream mixture-of-transformers couples video and action experts through shared attention, allowing future-scene prediction and trajectory generation to inform each other. At deployment, the model runs action-centrically and outputs only a trajectory. To ensure trackability, actions are parameterized as degree-7 Bernstein polynomials: the current state constrains the initial control points, and the network predicts five free control points, yielding C^2-continuous references with closed-form velocity, acceleration and jerk. FlowPilot is trained on a three-level depth pyramid spanning high-throughput simulation, photorealistic simulation, and real onboard data. In closed-loop simulation, it outperforms learning- and optimization-based baselines under increasing clutter and commanded speeds up to 8m/s. On a physical quadrotor, the full perception-to-action pipeline runs in under 18ms on a Jetson Orin NX and reaches 5.5m/s in cluttered indoor and forest environments using only onboard sensing and computation.
Chinese Translation
我们提出了FlowPilot，这是一种紧凑的世界动作模型，用于基于深度信息的实时机载无人机导航。与需要局部重建的地图-再优化流程或缺乏明确场景预测的端到端策略不同，FlowPilot通过流匹配共同去噪未来的深度观测和可执行轨迹。双流混合变换器通过共享注意力将视频和动作专家结合在一起，使未来场景预测和轨迹生成相互影响。在部署时，该模型以动作为中心运行，仅输出一条轨迹。为了确保可追踪性，动作被参数化为7次伯恩斯坦多项式：当前状态约束初始控制点，网络预测五个自由控制点，从而生成具有封闭形式的速度、加速度和冲击的C^2连续参考。FlowPilot在一个涵盖高吞吐量仿真、逼真仿真和真实机载数据的三层深度金字塔上进行训练。在闭环仿真中，它在日益复杂的环境和高达8m/s的指令速度下优于基于学习和优化的基线。在一台物理四旋翼上，完整的感知到动作管道在Jetson Orin NX上运行时间不到18毫秒，并在仅使用机载传感和计算的情况下，在复杂的室内和森林环境中达到5.5m/s的速度。
Human assistance in robotics spans around several tasks such as navigation, object manipulation, and placement, where a key challenge is selecting target destinations that align with human intentions or preferences. We focus on this challenge in the context of Virtual Placement (VP), the task of identifying all plausible target locations given scene context and human-centric constraints. This differs from traditional placement tasks that typically focus on a single, predefined target location. The VP problem is complex, as it requires both global and local reasoning about the scene's geometry, semantics, and plausibility. To address this gap, we introduce {\bf Assistant Placement Aria}, the first benchmark to explore diverse aspects of VP, including global, local, and human-centric constraints. It contains both synthetic and real indoor scenes annotated for three tasks: (i)~2D Panel Placement, (ii)~Sitting Suggestion, and (iii)~TV Placement. Each scene includes 2D images, a 3D point cloud, and a textual description of the objects within the scene. By contributing this benchmark, we aim to encourage further research in this underexplored and challenging field that is critically dependent on relevant data. We also evaluate several foundation models for object detection and segmentation on our benchmark.
Learning-based controllers can deliver exoskeleton assistance after training entirely in physics-based simulation, yet few controllers that address human-device co-adaptation have been validated on real users by whole-body metabolic measurement, the standard benchmark for assistive walking. Co-adaptation is challenging: as the device alters joint dynamics, the wearer reorganizes neuromuscular coordination, producing a non-stationary learning problem. Staged Multi-Agent Training (SMAT), a four-stage curriculum that progressively trains a musculoskeletal human actor and a bilateral hip exoskeleton actor, was introduced and shown to reduce simulated hip-muscle activation and provide positive assistance on hardware. This article provides the first physiological validation of SMAT. The policy was deployed on a hip exoskeleton and tested with eight healthy adults, with metabolic cost measured by indirect calorimetry across no-exoskeleton, passive, and active conditions. Active assistance lowered net metabolic rate by 19.7% relative to the passive device (p < 0.001). Biomechanical analysis confirmed predominantly positive hip mechanical power across all subjects (positive-power ratio 0.98), and the policy generalized across walking speeds and terrains. Together, these results show that a single simulation-trained SMAT policy, deployed without subject-specific retraining, delivers a significant metabolic benefit on real users while remaining robust beyond the conditions it was trained on.
Drones in human environments often lack spatially grounded in- terfaces for situated communication. We present OmniAI, an em- bodied aerial agent that supports surface-adaptive interaction by switching projection between an onboard screen and nearby en- vironmental surfaces. A servo-actuated MEMS laser projector renders text-and-image responses from a web-augmented LLM pipeline. Projection surfaces are detected online using RGB-D sensing and RANSAC plane fitting, without pre-mapped geometry. OmniAI provides functionally equivalent voice and gesture con- trol for both drone motion and projected content. By combining speech, mid-air gestures, adaptive projection, and aerial mobility, OmniAI demonstrates a mobile spatial AR interface for context- aware human-drone interaction.
SelfWAM: A Self-Grounded Unified World Action Model for Fast Robot Control
SelfWAM：一种自我基础的统一世界动作模型用于快速机器人控制
Pan, Bikang, Liu, Fan, Lu, Haotao, Wang, Jingya, Shi, Ye
Abstract
World Action Models (WAMs) improve robot policy learning by jointly modeling actions and future observations. However, conditioning future prediction only on the task prompt and observation context risks capturing generic task progression rather than the action-specific consequences of the executed action. We introduce SelfWAM, a unified self-grounded WAM built on a modality-specialized Mixture-of-Transformers (MoT) architecture that jointly predicts actions, action-conditioned future RGB frames, and robot self-masks, thereby grounding future prediction in the robot's visible body and its action-induced motion. During joint training, SelfWAM allows future visual queries to attend to a clean copy of the demonstrated action, turning the video branch into an action-specific consequence model while leaving the fast action-only inference path unchanged. To focus video learning on action-relevant visual changes, we use prompt-specific objectives for future robot self-mask prediction, which removes appearance details and provides a target whose temporal evolution is tightly coupled with the conditioning action. Together, clean-action conditioning and future self-mask supervision make future predictions more directly reflect how the executed action changes the robot's visible motion and the surrounding scene. Experiments on RoboTwin 2.0 and real-world manipulation tasks show that SelfWAM produces more action-sensitive futures and preserves fast policy inference, while improving policy performance.
Viscous stains, characterized by high viscosity and complex rheological properties, remain a major challenge for robotic surface cleaning. Conventional wiping often spreads the stain, while scrubbing provides stronger friction but risks damaging the surface. In this paper, we propose Push-Wiper, a framework that reformulates viscous stain cleaning as an aggregation problem. Push-Wiper employs a sponge to progressively gather stains through segmented pushing trajectories, followed by a post-processing phase that detaches the aggregated material and enables sponge self-cleaning. We adopt a stepwise strategy for stain gathering and leverage Diffusion Policy to generate adaptive pushing action sequences. These sequences are executed through our Arbitrary Surface Pose Interpolator (ASPI) and a hybrid force-position controller, allowing the method to generalize to stains with diverse spatial distributions. Push-Wiper achieves a cleaning score (CS), defined as the percentage of stain area removed, up to 130% higher than baseline methods. Without additional training, Push-Wiper also transfers in a zero-shot manner to solid residues, liquid spills, unseen viscous stains, and curved surfaces with varying geometries. Our experiments demonstrate the cleaning effectiveness of Push-Wiper and its strong generalization ability. The project website is available at https://push-wiper.github.io/.
When Prompts Control Robots: Prompt Injection Attacks in Multi-Agent Robotic Systems
当提示控制机器人：多智能体机器人系统中的提示注入攻击
Nagaraja, Neha, Bagari, Amisha, Bahsi, Hayretdin
Abstract
Large language models are increasingly integrated into autonomous robotic systems for task planning and control, but this integration exposes them to prompt injection attacks that can lead to unsafe decisions and physical harm. Multi-agent settings increase the risks through cross-agent contamination and broader attack surfaces. In this paper, we evaluate prompt injection attacks against an LLM-based multi-agent robotic system, considering both direct injections into task instructions and indirect injections through perception modules. In our experiments across varying attack-goal complexities and injection strategies in both single-agent and multi-agent settings, we show that prompt injection can induce adversarial actions while reducing task completion. We find that attacks can propagate from one agent to others through shared prompt structures, with impacts varying depending on prompt composition and the targeted agent. We further analyze how architectural changes affect LLM queries and, consequently, the attack success. To the best of our knowledge, this is the first study that systematically investigates prompt injection attacks in a multi-agent LLM-based robotic system.
ORCESTRA is a mixed-reality system for programming robot digital twins through no-code waypoint teaching and language-guided control. In a passthrough mixed-reality workspace, users place robot twins on real surfaces, teach trajectories, save robot-relative episodes, or issue spoken/typed commands that a vision-language model converts into structured digital-twin plans. Both interaction modes share a backend for metric grounding, embodiment-aware validation, preview, confirmation, and digital-twin execution. The system supports heterogeneous robot embodiments, including fixed-base manipulators, a mobile base, and a humanoid robot, demonstrating MR validation as a safety layer for language-guided robot programming before physical deployment.
SIPTraj: Map-Free End-to-End Trajectory Prediction via Physics-Guided Scene Interaction
SIPTraj：通过物理引导的场景交互实现无地图的端到端轨迹预测
Liu, Feifei, Wei, Zejun, Wang, Haozhe, Ye, Yazhi, Zhang, Yuying, Cheng, Jintao, Vong, Chi Man, Chen, Xieyuanli, Tang, Xiaoyu
Abstract
Trajectory prediction of surrounding agents is a prerequisite for safe planning and decision making in autonomous driving. Without high-definition (HD) maps, sensor-derived bird's-eye-view (BEV) features provide no explicit lane topology or drivable-area priors, making it inherently difficult to ground each agent in its surrounding scene context. Moreover, physical feasibility remains difficult to capture through data-driven learning alone, as kinematic constraints on agent motion cannot be explicitly encoded without structured supervision. Existing map-free predictors extract scene context in an agent-agnostic manner through a single fusion step and treat physical constraints only as output-level penalties, leaving both challenges unaddressed. We propose SIPTraj, a map-free trajectory prediction framework that jointly addresses scene grounding and physical feasibility. SIPTraj introduces a Hierarchical Agent-Scene Encoder (HASE) progressively grounding each agent in agent-guided scene evidence and refining inter-agent relations within the scene-grounded space. To tackle physical infeasibility in predicted trajectories, we develop a Physics-Guided Iterative Decoder (PGID). It conditions decoding on instantaneous kinematic states, propagating physical supervision into internal representations rather than output trajectories alone. Extensive experiments on nuScenes and Argoverse 2 Sensor show that SIPTraj surpasses prior map-free predictors and strong map-based baselines without any HD map at inference. Our code will be released as open-source.
Safe navigation under uncertain time-dependent blockage requires anticipating observations before committing to motion. We present StochSIPP, an exact contingent planner for temporal roadmaps with uncertain edge and vertex statuses revealed locally during execution. StochSIPP uses SIPP to generate certified-safe macro-actions that terminate at the next observation or the goal, and bounded AND/OR search over a cached action--observation graph to select actions for every reachable observation outcome. Optimistic and robust SIPP relaxations provide admissible lower and upper bounds for bounded AND/OR search. When every interval declared deterministically safe is truly safe, sensing is exact, and execution follows the planned timing, the resulting policy is provably collision-free. With correct independent probabilities and complete action and outcome generation, it minimizes expected arrival time within the roadmap and horizon. Experiments on controlled roadmap instances show that StochSIPP preserves the observed success of safe fixed-path baselines while reducing arrival time, and solves gated scenarios in which conservative fixed-path planners return no plan. A scalability study further reveals rapid growth as the number of simultaneously observed uncertain statuses increases.
Dynamic manipulation requires robots to infer target motion and respond promptly, yet existing World-Action Models (WAMs) typically condition only on the current frame and execute large backbones synchronously, limiting motion awareness and responsive control in dynamic scenes. We propose DynamicWAM, a compact WAM for dynamic object manipulation with dual-path motion conditioning. DynamicWAM introduces history-flow conditioning, encoding temporally aligned optical-flow frames alongside the current observation through a frozen pretrained video VAE to preserve spatial motion structure, while injecting kinematic descriptors of displacement, duration, velocity, and acceleration into the action expert to provide motion magnitude and timing. The two complementary paths are fused through joint world-action attention. A distilled compact backbone and real-time chunking (RTC)-based asynchronous execution further enable responsive control. On DOMINO, DynamicWAM achieves a 38.2% success rate and a 53.2 manipulation score, outperforming all evaluated baselines. Across 12 real-world tasks spanning linear, circular, and compound target motion, it achieves a 46.7% average success rate, exceeding the strongest baseline by 22.9 percentage points.
FastSAC-style methods significantly reduce humanoid motion training time but often suffer from notable performance degradation compared with PPO in whole-body tracking tasks. We target this speed-performance gap by introducing LooperMuscle, a composed expert policy learning framework that restores tracking quality while preserving high training efficiency. LooperMuscle combines a semantically structured mixture-of-experts actor, an expert-aware distributional critic, and contribution-routed replay with deferred curriculum scheduling. These three components form a closed training loop in which expert contributions guide data routing, routed data shape value learning, and value gradients in turn refine expert specialization. Empirically, our approach substantially outperforms vanilla FastSAC in motion tracking accuracy while requiring far less wall-clock time than PPO: where FastSAC trains in about 15 minutes but underperforms, and PPO achieves stronger results but requires about 6 hours, LooperMuscle recovers a substantial fraction of the remaining gap to PPO in roughly 45 minutes of simulation training, delivering practical efficiency for rapid policy iteration. The code will be released to benefit the research community at https://loopermuscle.github.io/.
We present Riemannian Informed Trees (RIT*), a planning framework that replaces Euclidean primitives in batch-informed search with their Riemannian counterparts. RIT* constructs a tighter, cost-consistent informed set, performs a nearest-neighbour search under an anisotropic distance metric, and evaluates edge costs efficiently via a cascading scheme. We further introduce a Collision-Adaptive Metric Refinement (CARM), which learns an obstacle-proximity cost field online from collision feedback, reducing the reliance on prior metric design in practical settings. Experiments across environments from 2-D to 14-D show that RIT* is competitive in low-dimensional and spatially constant-metric settings and produces substantially lower-cost solutions when the metric varies spatially in high-dimensional configuration spaces. Performance gains scale with anisotropy and dimension, reaching up to 13.0% improvement in median initial cost over BIT* in the 3-D anisotropic benchmark, up to 9.0% in median final cost over BIT* in 6-DOF manipulation, and 24.8-63.5% in a 14-DOF bimanual planning problem, where Euclidean-informed baselines degrade. Videos and code can be found here: https://muhayyuddin.github.io/ritstar/
GeminiPainter's sequence-formed pipeline comprised of perception, cognition, planning, and action stages
GeminiPainter的由感知、认知、规划和行动阶段组成的序列化管道
Cabrera, Miguel Altamirano, Fedoseev, Aleksey, Zhura, Iana, Tsetserukou, Dzmitry
Abstract
We present an autonomous robotic portrait-generation system combining real-time face detection, AI-based sketch generation, and robotic drawing. The system captures video frames, extracts facial regions, converts them into minimalist single-line sketches using the Gemini Vision API, optimizes stroke order through graph-based path planning, and executes smooth trajectories on a 6-DoF collaborative manipulator. This perception-cognition-action pipeline integrates computer vision, neural artistic abstraction, motion optimization, and robot control. User ratings on a 5-point scale were high for sketch quality 4.33, perceived execution 4.53, and user experience 4.65, indicating recognizable, appealing, and engaging robotic portraits.
Microrobots hold significant potential for various applications, where targeted navigation is a basic requirement. Deep reinforcement learning (DRL) has recently emerged as a powerful paradigm for fully autonomous microrobot navigation. Yet, current DRL-based approaches pay limited attention to learning efficiency and effectiveness, requiring hours to days for model training. Consequently, this impedes both rapid practical deployment and parameter optimization. To address these challenges, we present a learning framework that enables effective microrobot navigation policies to be trained within minutes. In the proposed framework, we develop a fully vectorized simulator with more than 10,000 artificial vascular environments, parallelizing dynamics, LiDAR-inspired perception, and feasibility checks across thousands of environments to achieve roughly 190,000 transitions per second. To achieve effectiveness in the fast training, we propose a task-shaping-regularization (TSR) reward framework. The TSR framework accelerates convergence, improves final performance, reduces action variation by at least 33.7%, and increases obstacle clearance by at least 2.1% across all evaluated scenarios. Results show that the proposed learning framework reduces training time to under 10 minutes, while supporting zero-shot deployment across distinct microrobot types and navigation scenarios. Collectively, this framework can substantially shorten the design loop and accelerate the deployment of autonomous microrobots.
Bicycle robots are fast and energy efficient, but their simple mechanical design and their underactuated and non-holonomic dynamics make highly agile maneuvers difficult to achieve. Here, we use Reinforcement Learning (RL) to enable a bicycle robot to learn and compose a diverse repertoire of dynamic acrobatic stunts. Using different RL formulations such as waypoint following, pose reaching, twist tracking, guided tracking, and motion imitation, the robot acquires autonomous single and multi-table forward and lateral jumps, steerable jumps, front flips, kip-ups, kip-downs, driving, wheelies, bunny hops, and three-point turns. To coordinate these behaviors, we introduce an orchestrator that transitions between policies using state-dependent triggers, enabling robust long-horizon acrobatic stunts. We validate the approach on the Ultra Mobility Vehicle (UMV), a custom bicycle robot, in simulation and hardware. The robot repeatedly traverses tables up to 1 m high, performs more than 15 consecutive autonomous jumps while following waypoints, handles previously unseen multi-table configurations, executes continuous repertoires of kipups, jumps, flips, kip-downs, over more than 20 consecutive trials, and performs more than 10 consecutive autonomous and steerable repertoires of wheelies, lateral jumps, and single-wheel jump downs. These results demonstrate that RL can endow bicycle robots with levels of agility previously associated primarily with legged platforms while preserving the speed and efficiency of wheeled locomotion, establishing a foundation for bicycle acrobatics.
Stipple: Real-Time Incremental Gaussian Splatting with Visual-Inertial Tracking
Stipple：基于视觉惯性跟踪的实时增量高斯点云渲染
Northoff, Kilian, de Mayo, Mateo, Cremers, Daniel
Abstract
3D Gaussian Splatting (3DGS) provides efficient rendering of photo-realistic scenes, but its heavy preprocessing and training steps make it a poor fit for applications that require real-time reconstruction in robotics or XR. This capability is important since it allows immediate feedback and interaction with new environments. Visual-inertial odometry (VIO) and simultaneous localization and mapping (VI-SLAM) systems, on the other hand, specifically target these real-time applications, which makes them a good choice for integration with 3DGS. We propose a new method that tracks and reconstructs simultaneously in real-time by leveraging an efficient visual-inertial tracking system based on Basalt together with a novel incremental method built on top of Brush, an efficient Rust-based GPU-vendor-agnostic implementation of 3D Gaussian Splatting. We show that many of the heavy preprocessing and training steps of 3DGS can be replaced with a more efficient incremental training strategy that has direct access to the information generated by the visual-inertial tracking system. Furthermore, we propose and combine multiple practical improvements to increase the efficiency of the training pipeline and adapt it to run in real-time, parallel to the tracking thread. This work highlights the value of exploiting the complementary nature of SLAM and 3DGS, and how that can lead to promising results for real-time 3D reconstruction.
Off-road mobility requires autonomous mobile robots to generalize across heterogeneous vehicle fleets and continuously changing terrain conditions. Existing cross-vehicle adaptation approaches generally assume flat terrain, while terrain-aware kinodynamic models often require platform-specific data collection and retraining. To this end, we propose VertiAKD, a unified framework for transferring and adapting off-road kinodynamic knowledge across diverse vehicles on geometrically and semantically complex terrain simultaneously. VertiAKD learns a shared mobility representation that jointly encodes vehicle configurations, trajectory transitions, and local elevation and semantic terrain features. Given limited data from a novel vehicle operating on unseen terrain, VertiAKD identifies the most relevant mobility descriptors and transfers their knowledge to initialize a terrain-aware kinodynamic model via function encoders, which is then periodically refined online from streaming observations without gradient-based retraining. We evaluate VertiAKD in the Verti-Bench simulator, built on the Chrono multi-physics engine, and on five physical configurations of the Verti-4-Wheeler platform. With only one minute of new trajectory data and associated terrain features, VertiAKD reduces long-horizon prediction error by up to 34.52% over direct mobility descriptor transfer across diverse unseen vehicle configurations and 94.43% over competing baselines. We further demonstrate robust closed-loop trajectory tracking in both simulation and physical experiments, highlighting the effectiveness of terrain-aware cross-vehicle knowledge transfer for accurate modeling and reliable off-road navigation.
Existing 6-DoF grasp detectors typically rank grasp candidates by detector confidence. However, our analysis on GraspNet-1Billion shows that detector confidence is often poorly aligned with grasp quality, causing successful grasp candidates to be ranked too low during execution. Motivated by this observation, we formulate grasp candidate re-ranking as a separate task for frozen detectors, aiming to improve candidate ordering without changing the detector or its grasp candidates. We propose GraRe, which estimates grasp quality from candidate attributes, shell-stratified local geometry, and object context. Candidate attributes condition the local geometric and object-context representations, and a Transformer fuses all three feature types. The predicted quality is combined with detector confidence to produce the final ranking. Experiments on GraspNet-1Billion with three frozen detectors show consistent improvements, with gains of up to 13.60 points in Average AP. Real-robot experiments further demonstrate robust grasping in cluttered scenes. These results show that improving candidate ranking provides a practical way to enhance frozen 6-DoF grasp detectors.
Object-oriented aerial vision-and-language navigation (VLN) requires searching for a described target and landing on it precisely, under long-horizon and closed-loop control. Guided by a target-descriptive instruction during navigation, perceptual priorities dynamically evolve: early-stage exploration prioritizes low-frequency spatial layout, and then shifts to high-frequency target details. Existing VLN methods model the varying perceptual requirements across navigation stages with identical visual tokens, leading to interference from irrelevant objects and background clutter. To this end, we therefore formulate long-horizon aerial navigation as a frequencypreference shift from spatial structure to local detail and propose FreqNav, a lightweight frequency-routing adaptive perception framework. Under a fixed computational budget, FreqNav dynamically reallocates visual tokens across frequency components according to the current navigation stage. A Frequency Token Router selects stage-relevant visual representations from dual-view observations, while a Phase-dependent Grounding Module anchors visual evidence through explicit supervision. A Diffusion Transformer then predicts smooth trajectories for continuous control. Experiments show that FreqNav outperforms strong baselines while achieving approximately 3x faster inference. Real-world deployment further demonstrates its effectiveness, efficiency, and practical potential for long-horizon aerial autonomy.
RL Bootstrapping of OpenVLA-OFT for a Novel Robot Embodiment
基于强化学习的OpenVLA-OFT在新型机器人体现中的自举
Nurtdinov, Damir, Kornaev, Alexei, Maloletov, Alexander
Abstract
Adapting a pretrained vision-language-action (VLA) policy to a new robot usually assumes embodiment-specific demonstrations. This assumption is especially restrictive for custom robots whose morphology differs strongly from the manipulators seen in large robot datasets. We study a harder setting: zero-demo embodiment alignment of OpenVLA-OFT on a cable-driven parallel robot (CDPR) with a simple gripper and a previously unseen control interface. Instead of supervised fine-tuning, we use reinforcement learning in simulation with dense geometric rewards computed from simulator state. The training is performed in two stages: a PPO stage for directional motion primitives, followed by GRPO continuation from the PPO checkpoint with an expanded instruction space that includes object-conditioned commands. On the four shared directional instructions, the average held-out success rate improves from 34.25\% after PPO to 53.50\% after PPO$\rightarrow$GRPO, with especially large gains on \texttt{move left} and \texttt{move backward}. In the GRPO stage we additionally introduce \texttt{move to
Chinese Translation
将预训练的视觉-语言-动作（VLA）策略适应于新机器人通常假设存在特定于体现的演示。这一假设对于形态与大型机器人数据集中所见的操纵器差异较大的定制机器人尤其限制性。我们研究了一个更为困难的情境：在具有简单抓手和先前未见控制界面的电缆驱动并联机器人（CDPR）上进行零演示的体现对齐。我们采用强化学习而非监督微调，在模拟环境中使用从模拟器状态计算的密集几何奖励。训练分为两个阶段：首先是用于方向运动原语的PPO阶段，随后是基于PPO检查点的GRPO继续阶段，扩展了包含对象条件命令的指令空间。在四个共享的方向指令上，平均保留成功率从PPO后的34.25%提高到PPO$
ightarrow$GRPO后的53.50%，其中在 exttt{move left}和 exttt{move backward}上获得了特别大的提升。在GRPO阶段，我们还引入了针对八个目标对象的 exttt{move to
Kinematic models provide reliable motion constraints for odometry estimation in featureless environments, where exteroceptive sensing degrades and IMU integration drifts. Learning-based kinematic models can achieve more accurate odometry estimation than model-based methods by capturing nonlinear effects; however, most existing learning-based models are trained on a single embodiment and generalize poorly to new embodiments. This generalization is difficult because the meanings and structures of proprioceptive measurements vary across embodiments, including the number of joints and ground-contact elements (e.g., wheels, feet). To address this challenge, we propose KING, a Graph Neural Network (GNN)-based kinematic model that explicitly incorporates robot embodiments by representing them as a common graph. We show that wheel and leg kinematic models can be expressed by a unified representation, enabling a single model for both wheeled and legged robots. Trained on datasets spanning diverse embodiments, KING provides a unified representation of wheeled and legged kinematics and achieves high-accuracy odometry estimation in real environments. KING estimates accurate odometry using only an embodiment description (e.g., a URDF file) and on-board proprioception (encoders and an IMU) and can be adapted to new robot embodiments through few-shot learning with only one minute of data, avoiding retraining from scratch on a new dataset for each robot. The project page is available at: https://smrg-aist.github.io/king_project_page/
Robot Task and Motion Planning (TAMP) algorithms enable autonomous operation by incorporating the specific functions and constraints of end-effector tools, such as grippers or soldering irons, directly into the planning process. In this paper, we explore sampling-based TAMP algorithms specifically designed for a critical subset of devices whose unique properties make traditional planning methods ineffective. Visibility-based instruments, such as exteroceptive sensors, cameras, flashlights and directional antennas, are essential across a vast array of human activities. The unique properties of these devices, and particularly, their field-of-view, render many widely used heuristics and distance metrics less effective. We introduce two new sampling-based algorithms, VisPRM and VisRRT, designed to tackle visibility-based tasks. VisPRM employs a hierarchical decomposition of the environment, leveraging the concept of visibility integrity, to efficiently sample configurations with a clear line-of-sight to the target. A specialized Inverse Kinematics solver enables VisRRT to "glance" in the direction of the target at opportune moments, facilitating the rapid discovery of key configurations. We show that VisPRM and VisRRT achieve a higher success rate and faster runtimes compared to adaptations of RRT, PRM and VIR, through both simulated and physical experiments.
VLAGuard: A Framework for Evaluating and Mitigating Physical Attention Hijacking in Vision-Language-Action Robots within Wireless Sensor Networks
VLAGuard：在无线传感器网络中评估和缓解视觉-语言-行动机器人物理注意力劫持的框架
Yin, Dongfu, Zhang, Jinquan
Abstract
Deploying Vision-Language-Action (VLA) robots as mobile edge nodes within wireless sensor networks (WSNs) requires robust protection against physical adversarial threats. We present VLAGuard, a framework to assess and mitigate a critical vulnerability: policy-critical action-to-vision attention hijacking. We first introduce a stress-test module, Visuomotor Attention-guided Semantic Attack (VASA), using printable patches to severely distract the robot's action-conditioned cross-attention. To counter this, we propose Attention-Protective Fine-Tuning (APFT), a defense that stabilizes spatiotemporal attention and enforces geometric consistency with zero inference overhead. Evaluations across simulated and physical WSN-assisted smart environments demonstrate significant robustness gains. APFT reduces the OpenVLA failure rate from 100.0% to 25.9% in LIBERO simulations. Furthermore, across 2,000 real-world trials, APFT improves the average success rate from 23.0% to 67.4% under severe patch attacks. This highlights that protecting attention pathways is important for improving the robustness of VLA-driven edge nodes in sensor networks.
Diffusion-Based Body Schema Learning Enabling Abnormal-State Adaptation in Musculoskeletal Robots
基于扩散的身体模式学习促进肌肉骨骼机器人在异常状态下的适应
Kawaharazuka, Kento, Ikemoto, Shuhei
Abstract
Musculoskeletal robots require an internal body schema that remains consistent under a wide range of physical state changes, including abnormalities such as muscle rupture and actuator jamming. Conventional approaches based on autoencoders or variational autoencoders learn average behaviors by projecting sensor and actuator signals into a low-dimensional latent space; however, exploration within the latent space alone has limited capability to handle out-of-distribution or abnormal states that are not included in the training data. To address this limitation, this study proposes a diffusion-based framework for body schema learning in musculoskeletal robots. Unlike generative models that operate through low-dimensional latent spaces, diffusion models can directly and iteratively estimate physically consistent sensor and actuator values in the high-dimensional space through a denoising process, even under partial observations and constraints, without requiring retraining. By formulating body schema adaptation as a gradient-guided denoising process, the proposed method enables adaptive estimation of appropriate muscle lengths and muscle tensions even under abnormal conditions such as muscle rupture and actuator jamming. The validity of the proposed framework is verified through simulation experiments using a musculoskeletal robot model.
Vision-Language-Action (VLA) models have emerged as a prominent paradigm for end-to-end autonomous driving; however, their efficient deployment is severely constrained by high computational latency and exposure bias arising from sequential autoregressive decoding. Conversely, while specialized diffusion policies enable low-latency, parallel execution, training them from scratch typically yields narrow, single-task architectures that lack holistic visual-linguistic reasoning. Successfully transforming pre-trained autoregressive generalists into parallel diffusion models could combine multi-task cognitive intelligence with execution efficiency, yet this transition presents a formidable architectural challenge due to mismatched attention patterns (causal versus bidirectional) and divergent optimization objectives. To bridge this divide, we introduce WAM-Diff2, a multi-task discrete diffusion VLA framework powered by a three-stage hierarchical distillation strategy. By structuring the architectural shift through progressive block-wise adaptation, block-wise distillation, and model-wise cross-scale distillation, WAM-Diff2 preserves the underlying semantic foundations of the base model while accelerating inference. Extensive evaluations across driving understanding, perception, and planning benchmarks demonstrate that WAM-Diff2 effectively mitigates exposure bias and achieves performance parity with autoregressive baselines. Crucially, the autoregressive-to-diffusion transition yields a 2.8x decoding speedup, which scales to an ultimate 15.1x acceleration when combined with system-level optimizations including FlashInfer and CUDA Graphs.
We propose OC-VLA++, an extension of OC-VLA for viewpoint generalization under limited camera coverage. While OC-VLA grounds robot actions in the camera coordinate system to align action supervision with visual observations, camera-space grounding alone can still overfit to the few viewpoints observed during training. OC-VLA++ addresses this limitation by introducing geometry-guided paired-view supervision and an explicit cross-view action-equivariance objective. Given paired observations of the same manipulation scene from geometrically related viewpoints, the model is trained such that their camera-space predictions correspond to the same robot-frame action. This objective explicitly supervises how action predictions should transform across viewpoints, rather than relying solely on image-level augmentation. Experiments demonstrate substantial improvements in unseen-view generalization under limited camera coverage, with performance degrading more gracefully under increasing camera displacement. These results establish cross-view action equivariance as an effective complement to observation-centric action grounding for robust real-world deployment.
VespaSeg: A Resource-Aware Ground-then-Segment Pipeline for Referring Expression Segmentation
VespaSeg：一种资源感知的先定位后分割管道用于指称表达分割
Wickramasinghe, Savindu Dilshan
Abstract
Referring expression segmentation requires language conditioned localization and pixel-accurate masks, but monolithic models can be costly to deploy. We present VespaSeg, a modular pipeline that grounds a text query with a compact vision-language model and converts the predicted box to a mask with MobileSAM. We study Florence-2-base, Florence-2-large, and Moondream2 grounders together with targeted adaptation of the grounding and segmentation stages. Under a repository-specific RefCOCO validation protocol containing the first expression for each of 3,811 referenced-object records, the adapted Florence-2-base pipeline obtains 73.64 mean intersection over union (mIoU) and 84.60 precision at IoU 0.5. On an NVIDIA RTX 6000 Ada GPU it processes 22.8 cached-image queries per second with 2.20 GB mean allocated GPU memory. A matched 500-query comparison gives 73.73 mIoU for Florence-2-base and 72.82 for Florence-2-large, while the base model is 1.70 times faster and uses 1.17 GB less allocated memory. Ablations show that ground-truth-box adaptation raises MobileSAM mIoU from 82.22 to 86.61 and that reducing the Florence-2 output-token budget from 64 to 32 preserves accuracy. These results support compact, modular grounding and segmentation, while also exposing the need for evaluation on the complete standard RefCOCO expression splits and deployment hardware.
Chinese Translation
指称表达分割需要基于语言的定位和像素级的精确掩码，但单一模型的部署成本可能较高。我们提出了VespaSeg，这是一种模块化管道，它通过紧凑的视觉-语言模型对文本查询进行定位，并使用MobileSAM将预测的框转换为掩码。我们研究了Florence-2-base、Florence-2-large和Moondream2定位器，并针对定位和分割阶段进行了有针对性的适配。在一个特定于库的RefCOCO验证协议下，该协议包含3,811个参考对象记录中的第一个表达，适配后的Florence-2-base管道获得了73.64的平均交并比（mIoU）和84.60的IoU 0.5时的精度。在NVIDIA RTX 6000 Ada GPU上，它每秒处理22.8个缓存图像查询，平均分配的GPU内存为2.20 GB。与500个查询的匹配比较显示，Florence-2-base的mIoU为73.73，而Florence-2-large为72.82，同时基础模型的速度快1.70倍，且使用的分配内存少1.17 GB。消融实验表明，真实框适配将MobileSAM的mIoU从82.22提高到86.61，并且将Florence-2输出令牌预算从64减少到32能够保持准确性。这些结果支持紧凑的模块化定位和分割，同时也暴露了在完整标准RefCOCO表达拆分和部署硬件上进行评估的必要性。
Manipulation involving rigid-deformable interactions, such as hanging clothes or dressing humans, is common in daily life, making it essential for household robots. Compared to single-object manipulation or interactions between rigid bodies, these tasks are particularly challenging due to the rich multi-point contacts and the complex dynamics of the deformable bodies during interaction. Therefore, object-centric representations such as 6D poses or structural points without task-specific information become insufficient for these interactions. In this work, we propose a hybrid correspondence-based representation tailored for rigid-deformable interactions. First, to capture intricate interaction information, we introduce structure-, task-, and interaction-aware sparse keypoints. The keypoints are generated based on the global structures of both rigid and deformable objects, and filtered by their local interaction contacts. However, tracking these sparse keypoints through the interaction remains difficult due to the high-dimensional dynamics of deformable objects. Therefore, we further construct dense correspondences on the deformable objects for accurate keypoint tracking throughout the manipulation. This hybrid design combines the advantages of both representations: sparse keypoints encode rich, task-specific information for fine-grained manipulation, while dense correspondences ensure efficient tracking and generalization to novel deformations, shapes, and scenarios. Together, they enable one-shot transfer to new tasks with minimal demonstrations. Extensive experiments demonstrate the effectiveness and broad applicability of our method.
In contact-rich manipulation, visual observations primarily guide motion in free space, whereas tactile observations become particularly informative during contact. However, standard Transformer-based visuo-tactile policies typically rely on either token concatenation or learnable gating. These approaches lack explicit contact-aware priors, making it difficult to efficiently learn effective cross-modal representations from demonstrations. To address this limitation, we propose CAAT, a lightweight contact-aware framework that explicitly incorporates contact priors through attention scaling and dynamic tactile masking. Specifically, CAAT emphasizes visual information before contact and tactile information during contact. It also suppresses static background tokens by comparing the current tactile observation with a non-contact reference. CAAT can be integrated into commonly used Transformer-based policies without modifying their action decoders. In simulation, integrating CAAT with ACT improves the average success rate by 18.0 percentage points over direct visuo-tactile fusion and by 10.0 percentage points over gated fusion. In real-world experiments using a visuo-tactile UMI platform, CAAT achieves an average success rate of 60.0% across ACT, Diffusion Policy, and $\pi_0$, outperforming the strongest baseline by an average of 21.1 percentage points. These results demonstrate that explicit contact priors and dynamic tactile masking are effective in improving visuo-tactile policy learning and task performance of diverse policy architectures. https://mrjiangjm.github.io/caat/
Although recent robot perception research emphasizes training on data from diverse environments to improve generalization, most existing methods still rely on centralized learning, which is inefficient and difficult to scale across heterogeneous robot platforms. Federated learning (FL) offers an alternative by enabling distributed training without raw data transfer, but it suffers from severe performance degradation under domain shifts caused by heterogeneity across clients. In real robotic deployments, data distributions often overlap across platforms, environments, and sensing conditions, making it difficult to partition clients into clearly separated domains. However, this characteristic breaks the assumption of clearly separable client domains commonly used in clustered FL. To address this gap in robot perception, particularly in depth estimation, we introduce two realistic and unexplored non-IID scenarios that reflect heterogeneity in terms of platform, environment, and depth distribution. We then propose FeDepth, a descriptor-based clustered FL framework that models client relationships through soft clustering. Unlike hard clustering methods that assume clearly separated clusters, FeDepth allows clients to participate in multiple clusters, capturing continuous and ambiguous domain transitions commonly observed in robotic environments. Extensive experiments demonstrate that FeDepth consistently improves robustness over standard FL and clustered FL baselines across multiple depth estimation architectures, providing a practical and effective solution for federated robot perception. Our project page is available at https://vision3d-lab.github.io/fedepth/.
Complete Motion Planning using Workspace-Fibered Decomposition for nR-Planar Manipulator
基于工作空间纤维分解的 nR 平面操纵器完整运动规划
Rath, Aayush, Thomas, Antony
Abstract
We propose a workspace-fibered decomposition framework for motion planning in nR planar redundant manipulators operating in cluttered environments. Rather than planning directly in the full n-dimensional configuration space, the method incrementally constructs obstacle-constrained reachable workspaces of lower-dimensional non-redundant sub-chains and recursively lifts them through redundant orientation fibers. This yields a sequence of reduced planning manifolds that preserve branch-consistent reachability structure while avoiding explicit construction of the full configuration-space obstacle geometry. We first establish that, for planar position-only manipulators, the obstacle-constrained reachable workspace induced by the minimal non-redundant sub-chain provides an exact characterization of feasibility with respect to the connected component of the start configuration, enabling early infeasibility detection prior to introducing redundant degrees of freedom (DOF). We then introduce an incremental fiber-lifting procedure that propagates reachable workspace structure through successive redundant links while enforcing local inverse-kinematic branch consistency using Jacobian determinant continuity constraints. The resulting representation admits efficient reduced-space planning directly on recursively-constructed workspace-fiber manifolds. Experimental results on redundant nR planar manipulators demonstrate that the proposed construction preserves collision-free connectivity structure across successive lifting stages while substantially reducing collision checking complexity relative to direct configuration space reasoning.
Chinese Translation
我们提出了一种工作空间纤维分解框架，用于在拥挤环境中进行 nR 平面冗余操纵器的运动规划。该方法不是直接在完整的 n 维配置空间中进行规划，而是逐步构建低维非冗余子链的障碍约束可达工作空间，并通过冗余方向纤维递归提升它们。这产生了一系列减少的规划流形，保留了分支一致的可达性结构，同时避免了对完整配置空间障碍几何形状的显式构建。我们首先证明，对于仅平面位置的操纵器，由最小非冗余子链诱导的障碍约束可达工作空间提供了关于起始配置连通分量的可行性精确表征，从而在引入冗余自由度（DOF）之前实现早期不可行性检测。然后，我们引入了一种增量纤维提升过程，通过连续的冗余链传递可达工作空间结构，同时使用雅可比行列式连续性约束来强制局部逆运动学分支一致性。最终的表示允许在递归构建的工作空间纤维流形上直接进行高效的减小空间规划。对冗余 nR 平面操纵器的实验结果表明，所提出的构建在连续提升阶段保持了无碰撞连通结构，同时相对于直接配置空间推理显著降低了碰撞检测的复杂性。
Havrylov, Volodymyr, Janjoš, Faris, Look, Andreas, Mathes, Jürgen, Geiger, Andreas
Abstract
End-to-end autonomous driving (E2E AD) systems integrate perception, prediction, and planning into a single differentiable architecture. While these models show great promise, their standard training often relies on output-only supervision, which can lead to weak gradients for the hidden layers of increasingly complex models. Recent works have integrated vision-language model (VLM) supervision for latent features to address this, yielding substantial empirical gains, yet leaving the underlying theoretical mechanisms poorly understood. Our investigation into this methodology reveals that the resulting performance gains stem not from VLM reasoning capabilities, as previously assumed, but rather from the latent connections forged between the E2E AD model and ground-truth (GT) data during training. Building on this insight, we propose a probabilistic deep supervision framework that regularizes intermediate latent representations directly from GT data. By treating model latents as reparameterizable distributions, we optimize the architecture via the Evidence Lower Bound (ELBO). Our evaluations conducted on the nuScenes dataset demonstrate that supervising trajectory-related latents with future GT paths consistently improves planning performance. Using identical training data and E2E architectures, our method achieves an 8% reduction in planning L2 error and a 3% decrease in collision rates compared to competitive vectorized baselines, all while incurring negligible computational overhead.
Autonomous endoscopic navigation can reduce clinicians' operational burden, yet robust control remains challenging due to tissue deformation, transient occlusions, and rapidly changing viewpoints. Existing learning-based policies typically predict actions from current observations without explicitly modeling future dynamics, limiting their robustness and reliability in safety-critical settings. World Action Models (WAMs) offer a promising alternative by coupling predictive visual dynamics with action generation, but extending them to robotic endoscopy remains challenging due to limited training data, restricted viewpoint diversity, deformable anatomy, and high inference latency. We present EndoWAM, which is, to our knowledge, the first WAM for generalizable robotic endoscopic navigation. EndoWAM introduces future grounding, which predicts task-relevant target regions in future observations from intermediate denoising features of a video world model. Specifically, EndoWAM couples a lightweight diffusion transformer for future target-region prediction with a discrete action expert through a shared predictive representation. This design injects target-aware supervision into predictive dynamics modeling, improving robustness to visual degradation and viewpoint changes while enabling real-time control in a single denoising pass. We further introduce EndoMotion, a robotic endoscopic motion dataset spanning three anatomically distinct procedures: ureteroscopy, esophagoscopy, and endoscopic retrograde cholangiopancreatography (ERCP). EndoWAM consistently outperforms all baselines and alternative grounding strategies, while demonstrating strong zero-shot generalization to unseen viewpoints, environments, and targets. These results establish EndoWAM as a predictive, target-grounded framework for accurate, generalizable, and long-horizon navigation in visually constrained endoscopic environments.
Despite recent progress in Vision-Language-Action (VLA) models for robotic manipulation, the action chunk remains a weakly structured interface. Existing work typically flatten each chunk into per-timestep controls, relying on implicit data learning that manifests as jagged motion and boundary discontinuities during physical execution. To address these limitations, we introduce Hermite trajectory priors, parameterizing the chunk trajectory as a piecewise cubic Hermite curve defined by endpoint positions and velocities to explicitly enforce smoothness and continuity. We instantiate this fixed operator across discrete autoregressive and continuous generative paradigms via three variants: (1) Hermite Tokens, which predict quantized boundary variables autoregressively; (2) Hermite Scaffold, which decomposes clean actions into a base scaffold and residuals; and (3) Hermite Regularization, which applies the prior strictly as an auxiliary training objective. Across simulation benchmarks and real-robot platforms, Hermite Regularization achieves superior performance among these three variants, improving {\pi}0.5 baseline success rates from 95.9% to 98.7% on LIBERO, 85.7% to 90.9% on LIBERO-plus, and 63.4% to 90.0% across four real-robot tasks without additional inference overhead. Trajectory analyses reveal that explicitly structuring trajectory priors serves most effectively as a learning inductive bias rather than a runtime constraint.
Mobile manipulation requires a robot to coordinate base and arm motion under continuously changing viewpoints and contact conditions, within an action space far larger than that of fixed-base manipulation. Existing Vision-Language-Action (VLA) policies are limited in two respects. (i)They map observations directly to whole-body action chunks, searching this large action space without an explicit task-space motion plan, which makes coordinated base--arm prediction imprecise. (ii)They execute the predicted chunk open-loop, without checking whether the actions can realize the motion the policy intended, so control errors and unmodeled contacts accumulate into a gap between planned and realized motion. We present DreamTrajectory, a trajectory-guided framework for language-conditioned mobile manipulation that introduces one component for each limitation. Addressing(i), DreamTrajectory jointly predicts an intention-level end-effector trajectory and a whole-body action chunk in a single action expert, so that the trajectory explicitly guides base--arm action generation instead of remaining implicit. Addressing(ii), a lightweight trajectory world model predicts the trajectory that a candidate action chunk would induce, and a test-time search--predict--score procedure selects the candidate best aligned with the planned trajectory. On MS-HAB, trajectory guidance raises average success from 32.3% to 47.5% and test-time refinement further to 54.8%, with the largest gains on contact-rich articulated-object tasks. On three real-world mobile manipulation tasks, the corresponding average success rates are 63.3%, 81.7%, and 90.0%.
SG-WAM: Self-Guided World Modeling in Geometry-Aware Policy Space
SG-WAM：自指导的几何感知策略空间中的世界建模
Zhao, Ruiteng, Zhang, Zhengshen, Su, Yue, Wang, Wenshuo, Li, Jiahui, Yang, Zhiyuan, Tay, Francis E. H., Ang Jr., Marcelo H., Zhu, Haiyue
Abstract
World Action Models (WAMs) couple action generation with prediction of future states. Their effectiveness depends on whether future dynamics are modeled in a space that is both aligned with action generation and sufficiently geometry-aware to capture where and how actions change the scene. Existing WAMs typically satisfy only part of this requirement, relying on either perceptually heavy observation-space targets or auxiliary latent spaces that are not jointly structured for action relevance and geometry. We propose SG-WAM, a self-guided framework that learns geometry-aware action-conditioned dynamics directly in the policy-derived representation space. SG-WAM introduces learnable dynamics tokens and a Self-Guided World Predictor that forecasts their future latent states conditioned on intervening robot actions. Prediction targets are generated by an exponential moving average copy of the same policy backbone, providing stable supervision within the representation family used by the action expert. Geometric supervision further structures the policy image-token representations, providing spatially grounded context for the dynamics tokens and yielding a future-alignment space that is both action-relevant and geometry-aware. Latent future prediction, geometric grounding, and flow-matching action generation are jointly optimized end-to-end in a unified framework. Built on a 0.9B model without large-scale embodied pretraining, SG-WAM achieves 98.5% average success on LIBERO and 73% on LIBERO-Plus, while outperforming strong baselines in both in-distribution and out-of-distribution real-world evaluations.
We address the problem of understanding when and why Vision-Language-Action models struggle with contact-rich manipulation tasks that require precise physical interaction. Prior work has primarily focused on addressing contact failures through force-augmented architectures and training-time regularizers, yet the root causes of these failures remain underexplored. We identify two distinct failure modes underlying this gap. Precision failures are rooted in a flow-matching policy training mismatch, and force failures arise from the distinctive structure of force signals. We address each failure mode with a targeted mechanism and combine them into FACT, which achieves 66% average success rate across five contact-rich tasks against 41% for the best prior baseline, in an evaluation spanning almost 2,500 real-world rollouts.
General-purpose humanoid trackers can execute diverse references, but their zero-shot coverage depends on large embodied corpora that are costly to extend. Text-to-motion generators offer scalable supervision, yet models trained on human motion or retargeted data inherit a gap between kinematic plausibility and robot executability. Existing one-way pipelines fix either the generated corpus or the reward tracker. We introduce GenTrack, an online generator--tracker framework that alternates execution-grounded, group-relative generator alignment with tracker training on newly generated references; anchoring and rehearsal constrain drift. On Unitree G1, we evaluate GenTrack with ProtoMotions and SONIC backbones across three zero-shot tracking splits including public AMASS and LAFAN benchmarks, and a private out-of-distribution test set of 1,024 prompt-motion pairs in the wild. The online co-training strategy consistently produces generators that output more robot-executable motions with strong semantic alignment, and trackers with markedly broader zero-shot coverage and improved tracking accuracy, especially on out-of-distribution references. These results demonstrate that joint online post-training effectively narrows the executability gap between retargeted references and robot-native motion, advancing zero-shot humanoid control without additional data collection and beyond the limitations of a static reference pool.
When Replanning Becomes the Bottleneck: Budgeted Replanning for Embodied Agents
当重新规划成为瓶颈：面向具身体代理的预算重新规划
Liu, Shuaijun, You, Feiyang, Chen, Xingwei, Su, Ningxin
Abstract
Embodied agents replan frequently to recover from execution drift, partial observability, and coordination hazards, but each LLM-based replanning call can consume an accumulated textual context that grows over time and across agents. Once this context becomes large, replanning latency develops heavy tails and can miss real-time deadlines even when task success remains high, a failure mode that is hard to detect from average latency or success alone. We present BRACE, a controller that formulates replanning as a budgeted control loop by deciding whether to replan, selecting a replanning mode, and allocating an explicit token budget and latency service-level objective (SLO) while accounting for optional efficiency modules. As a reusable component, we introduce E-RECAP, a cost-aware progressive token pruning method that predicts token utility and prunes replanning contexts across transformer layers while preserving critical head and tail tokens. Across Meta Habitat, RoboFactory, and AirSim, BRACE with E-RECAP reduces replanning-call token counts by 62-92% and SLO violation rates from 85.5-100.0% to 4.7-50.0% in settings where task success is already saturated. In a harder RoboFactory setting where open-loop, frozen-plan, and No BRACE all fail, BRACE + E-RECAP reaches 80.0% success with 4.6% SLO violations, demonstrating that tail-aware per-call budgeting is effective across embodied platforms.
Dynamic manipulation is a critical capability for robots operating in complex and dynamic environments, where robots must interact with objects that are moving or require rapid adjustments. However, learning models for dynamic manipulation tasks face two major challenges: (1) the combinatorial complexity of dynamic scenarios leads to substantial data requirements, and (2) rapid variations in dynamics require real-time and accurate policy execution. In this paper, we propose DynamicManip to address these challenges through an efficient data augmentation pipeline and a low-latency imitation policy. We first propose a static-to-dynamic augmentation pipeline that synthesizes diverse dynamic manipulation demonstrations from a single static demonstration. Second, we introduce a dynamic-aware adaptive policy that adjusts its inference frequency according to task dynamics, enabling responsive and effective dynamic manipulation. Third, we build a dynamic manipulation benchmark, which includes diverse dynamic tasks with an automatic evaluation system for scalable and consistent assessment. Extensive experiments in both simulation and the real world demonstrate that DynamicManip not only provides significant improvements in data efficiency but also achieves better performance in dynamic manipulation tasks, with a mean success rate 18.4 percentage points higher and policy-query latency 32.9% lower.
Rapid Embodiment Adaptation for Quadrupedal Locomotion
四足运动的快速体现适应
Li, Dichen, Ai, Bo, Bohlinger, Nico, Peters, Jan, Su, Hao, Christensen, Henrik I.
Abstract
Humans readily adapt their movements as their bodies change through aging, injury, or load carrying, but learning-based robot policies often break when hardware properties shift. We introduce an online embodiment adaptation framework for quadrupedal locomotion that infers embodiment parameters from short interaction histories and conditions control on the inferred hardware state. Our method pairs a generalist policy trained under embodiment randomization with a lightweight adaptation module that identifies physical changes within half a second. We evaluate two representative forms of embodiment variation: joint-range constraints and trunk-mass changes, corresponding to joint-level kinematic degradation and body-level dynamic variation. In simulation, the module accurately estimates these changes and enables closed-loop control that substantially outperforms policies conditioned directly on interaction history. On a real Unitree Go2 robot, our system maintains stable locomotion under severe instances of the evaluated changes, including a fully locked leg and a 5 kg payload, where non-adaptive methods fail. These results demonstrate the practicality of explicit online embodiment identification for rapid adaptation to joint-limit and payload-mass changes, and provide a step toward handling broader forms of uncertain, degraded, or changing robot hardware.
Learning Smooth SE(3) Trajectories under Left-Invariant Riemannian Metrics
在左不变黎曼度量下学习平滑的 SE(3) 轨迹
Wu, Yuwei, Kumar, Vijay
Abstract
Optimal trajectory generation for rigid-body motions on Lie groups can be formulated as a variational problem that minimizes energy functionals defined by Riemannian metrics. While closed-form solutions exist for special cases such as product metrics and rest-to-rest boundary conditions, solving the general problem with arbitrary boundary states and coupled rotational-translational metrics often requires computationally expensive numerical boundary value solvers. These limitations restrict the use of geometrically consistent trajectory generation in real-time robotic planning and control. This paper presents a learning-based framework for approximating higher-order smooth trajectories on SE(3) under general left-invariant Riemannian metrics. The method parameterizes body-twist trajectories using high-order polynomials and relies on a neural network to learn a subset of the polynomial coefficients and the trajectory duration. The remaining coefficients are analytically determined to enforce the boundary conditions. The training of the network is guided by losses derived from Euler-Lagrange optimality conditions, metric-weighted smoothness objectives, and feasibility constraints. The metric-conditioned framework enables generalization across diverse metric structures and motion conditions. Extensive numerical experiments demonstrate that the proposed approach generates smooth trajectories that closely approximate solutions from numerical optimization while achieving millisecond-level inference times. We demonstrate two practical applications of the proposed framework: real-time generation of diverse motion primitives with waypoint traversal, and refinement for quadrotor flight under dynamic conditions. These results suggest that learning-based motions with geometric structure can provide an efficient alternative to conventional optimization-based methods for trajectory generation on SE(3).
Recent Vision-Language-Action (VLA) models achieve promising performance in robotic manipulation, typically measured by success rates aggregated over predefined object configurations, an evaluation that implicitly assumes spatially uniform competence across the workspace. However, this assumption does not hold: even with the instruction and every other scene factor held fixed, merely relocating a task-irrelevant distractor can sharply raise the failure probability within localized, spatially coherent regions, which we term Positional Blind Spots (PBS). In this paper, we propose a two-stage black-box framework to uncover and mitigate PBS. During the uncovering stage, we grid the workspace and apply a one-sided log-likelihood-ratio test to localize PBS cells with significantly elevated risk. During the mitigation stage, we fine-tune the policy via LoRA on demonstrations collected from these PBS regions, improving competence there while largely preserving performance across the rest of the workspace. We evaluate our framework on five state-of-the-art VLA policies across two benchmarks, and find that PBS are pervasive and spatially concentrated in all of them, with failure rates up to 0.58. Our search strategy achieves an average F1-score of 0.678, outperforming random search and adaptive sampling baselines by 0.268 and 0.178, respectively. Guided by the discovered regions, targeted fine-tuning reduces the overall failure rate by 40.00%--85.19%.
Perception-and-action system for humanoid robot task execution in construction
用于建筑任务执行的人形机器人感知与行动系统
Liu, Yanxi, Liu, Yizhi
Abstract
Humanoid robots, with their human-like shape and multi-tasking capabilities, are well-aligned with human-dominated workplaces, like those in civil and construction engineering, where they could collaborate with human workers or autonomously perform physically demanding and hazardous tasks. Despite this promise, limited research has explored how to endow these robots with the practical capabilities needed to perform construction tasks. To this end, this study proposes a novel perception-and-action system that enables humanoid robots to learn and perform construction tasks from worker demonstrations. This system contains two deep networks: Humanoid-PoseNet, which extracts human postures and translates them into mechanically feasible poses for a humanoid robot; and Humanoid-ActionNet, which learns robot-executable actions based on these translated poses. Experimental results demonstrate that the humanoid robot reliably executed eight construction-related actions, achieving an average motion-tracking error of 82.45 mm MPJPE (Mean Per Joint Position Error). This work provides an early step toward deploying humanoid collaborators in construction.
AffordTrajDP: Dynamic Affordance-Guided Visuomotor Policy Learning for Robotic Manipulation
AffordTrajDP：动态可供性引导的视觉运动策略学习用于机器人操作
Wu, Gaoyuan, Shan, Ziyu, Du, Haoyang, Jiang, Yuyao, Wang, Ziwei
Abstract
Affordance-guided imitation learning has shown impressive performance in robotic manipulation tasks by compressing visual perception into task-specific geometric constraints (e.g., fixed contact points). However, the commonly used static affordances can become inconsistent in precision-critical tasks or under object location perturbations, leading to post-contact trajectory drift. To address this issue, we propose AffordTrajDP, a dynamic framework that constructs affordance trajectories via object-centric temporal propagation to guide the progressive manipulation process. Specifically, given an RGB-D observation, our core insight is that a retrieved anchor affordance, which captures the desired contact point between the end-effector and the target object, can be propagated forward via affordance propagation, using the object's SE(3) pose as a natural propagation medium, to yield an affordance trajectory that provides temporally consistent, state-aware guidance throughout execution. AffordTrajDP achieves 70.0% average success rate on ManiSkill3, outperforming strong baselines by up to 17.8%. Real-world experiments on Galaxea A1 and UR7e robotic arms, covering StackCube, PickCup, AdapterInsertion, Ring-on-Peg, Put-in-Bowl, and USB Insertion, further validate robustness under object placement variations and appearance changes, with seen and unseen object instances evaluated on Galaxea A1, and ablations confirm the contribution of each proposed component.
This paper studies feedback Nash equilibrium (FBNE) seeking for multi-agent trajectory planning in nonlinear dynamical systems with unknown agents' objectives and state-dependent inter-agent coupling. While dynamic game theory provides a principled framework for such problems, existing approaches typically assume fully rational agents with known objectives or rely on fixed regularization, limiting their ability to capture bounded rationality and spatially varying interaction intensity in safety-critical settings. To this end, we propose a KL-regularized dynamic game with a state-dependent weight that adaptively balances optimality and behavioral priors. To infer unknown cost parameters from demonstrated behaviors, we develop a context-aware inverse game module based on maximum-entropy inverse reinforcement learning with physics-informed regularization, ensuring structural consistency with the forward game. We establish per-iteration well-posedness of the regularized local game and show that the adaptive weighting function remains Lipschitz continuous under bounded nominal-trajectory updates. Numerical simulations and multi-robot experiments on cooperative navigation and merging scenarios validate the effectiveness of the proposed framework.
LLM-based multi-agent coordination faces a fundamental trade-off between efficiency and adaptivity in dynamic environments. Existing approaches typically rely on repeated LLM invocations or multi-round communication to adapt decisions during execution, introducing substantial latency and making coordination vulnerable to asynchronous progress and environmental changes. Conversely, one-shot planning reduces coordination overhead but produces open-loop plans that can quickly become stale or fail when actions depend on other agents and the environment. We introduce SyncPlan, a plan-execute-correct framework for long-horizon coordination through explicit synchronization and adaptive correction. Given the state and team-level task, a centralized LLM coordinator generates per-agent action chains in a single planning call. During execution, explicit wait primitives and deadlock detection enforce inter-agent and agent-environment dependencies, while a lightweight Plan Staleness Detector continuously assesses the remaining plan and triggers replanning when environmental changes invalidate its assumptions. We further optimize the coordinator through SFT and planning-oriented RL with dense task progress and outcome-level execution feedback. Experiments on the public Overcooked benchmark and the complex Honor of Kings environment show that SyncPlan achieves state-of-the-art task success rates while using less than 0.05% of the wall-clock runtime compared with existing LLM-based coordinators. Code and datasets will be made publicly available.
Biological wet-lab protocols are written for trained researchers and often leave routine operations, state-dependent conditions, and contextual parameters implicit, making them difficult to translate into robot-executable actions. We present ProtoAct, a structured protocol-grounding framework that converts free-form biological procedures into state-aware, embodiment-ready action sequences. ProtoAct uses ProtoRAG to retrieve manually annotated examples for context-sensitive parsing, employs RefineChecker to detect and revise missing or inconsistent steps, and applies ActSchema to map the refined procedure into constrained JSON function sequences. We further introduce BioP2E, for which we manually annotate 22 cell-culture protocols into 258 monitoring conditions, 910 executable subtasks, and 962 grounded action calls. Evaluation across seven large language models demonstrates that ProtoAct can be effectively instantiated with different backbones. Ablations confirm that retrieval, posterior checking, and schema constraints make complementary contributions. The parsed subtasks further support demonstration collection and VLA model training, enabling successful execution in both simulation and real-robot settings. ProtoAct thus provides a practical interface between biological protocol understanding and embodied robotic execution.
Bridging the Sim-to-Real Gap in Parallel-Link Leg Mechanisms via Simulator-Side Dynamics Normalization
通过模拟器侧动力学归一化弥合并联连杆机构的仿真与现实差距
Hong, Jinsong, Kim, Jangho, Lee, Jihwan, Kim, Donghyun, Oh, Sehoon
Abstract
This paper addresses the sim-to-real gap in dynamics arising when a parallel-link mechanism is represented by a serial-tree surrogate in simulation. Conventional Jacobian-based state and torque mappings preserve consistency with the kinematic and virtual-work relations but do not account for the coordinate-induced redistribution of actuator inertia and damping and the linkage inertia omitted during serial-tree reduction. To address this gap, Simulator-Side System Normalization (S3N) is proposed to normalize the serial-tree simulator's effective dynamics while preserving its tree topology. S3N-Act incorporates actuator inertia and damping into the serial-coordinate dynamics through coordinate transformation, whereas S3N-Full restores residual linkage inertia by separately identifying actuator- and leg-level frequency responses. In the 2-DoF validation, S3N-Full reduced the joint-position and torque RMSEs by 80.9% and 82.1%, respectively, relative to the Jacobian-mapping baseline. During pitch-in-place motion, S3N-Act and S3N-Full reduced the RMSE of the ground reaction force norm by 65.1% and 62.4%, respectively. During circular locomotion, S3N-Full reduced the phase-averaged, command-normalized sim-to-real gap from 17.3% to 9.9%. These results show that simulator-side normalization improves motion- and force-level sim-to-real consistency. It enables policy training in a serial-tree framework with hardware-consistent dynamics that better represent the physical parallel-link mechanism.
Chinese Translation
本文探讨了在仿真中将并联连杆机制表示为串联树代理时出现的动力学仿真与现实之间的差距。传统的基于雅可比矩阵的状态和扭矩映射虽然保持了与运动学和虚功关系的一致性，但未考虑由于坐标引起的执行器惯性和阻尼的重新分配，以及在串联树简化过程中省略的连杆惯性。为了解决这一差距，提出了模拟器侧系统归一化（Simulator-Side System Normalization, S3N），旨在在保持树状拓扑结构的同时归一化串联树模拟器的有效动力学。S3N-Act通过坐标变换将执行器惯性和阻尼纳入串联坐标动力学，而S3N-Full则通过分别识别执行器和腿部的频率响应来恢复剩余的连杆惯性。在2自由度验证中，S3N-Full相对于雅可比映射基线将关节位置和扭矩的均方根误差（RMSE）分别降低了80.9%和82.1%。在原地俯仰运动中，S3N-Act和S3N-Full分别将地面反作用力范数的RMSE降低了65.1%和62.4%。在圆形运动中，S3N-Full将相位平均的命令归一化的仿真与现实差距从17.3%降低到9.9%。这些结果表明，模拟器侧归一化改善了运动和力级别的仿真与现实一致性。它使得在具有硬件一致性动力学的串联树框架中进行策略训练成为可能，更好地代表物理并联连杆机制。
Recent advances in robot learning for manipulation have increased the importance of collecting real-world demonstration data. However, existing robotic systems primarily focus on end-effector manipulation, making it difficult to teach and execute manipulation tasks involving body-surface contact with the arms and chest. This paper presents TWINS (Tactile Wearable Isomorphic Arm Networked System), a robotic system for manipulation involving body-surface contact. TWINS consists of a Wearable Dual-Arm Device, which is worn by the operator, and an Isomorphic Robot with the same joint configuration and external dimensions. Distributed tactile sensors embedded in the chest and arms enable the measurement of body-surface contact synchronized with joint motion. Using the Wearable Dual-Arm Device, we collected demonstrations for four manipulation tasks involving body-surface contact. We then trained imitation learning policies using the collected demonstrations and deployed them on the Isomorphic Robot, enabling manipulation guided by body-surface tactile observations. Experimental results demonstrate that TWINS provides a unified robotic system for demonstration, learning, and execution of manipulation involving body-surface contact. https://mmurooka.github.io/twins-project-page/
Maintaining a stable pose during physical interaction is a significant challenge for aerial robots, often limiting their use in contact-based tasks. This paper presents a novel uncrewed aerial vehicle (UAV) platform designed to transition from unconstrained flight to a stable, constrained work platform via environmental anchoring. Our system comprises: 1) a multirotor with a tilt-rotor mechanism that decouples pitch from forward motion, enabling stable hover at non-zero pitch angles, and 2) a novel underactuated, cable-driven, prismatic gripper featuring compliance to adapt to irregular geometries, designed to stabilize the UAV by anchoring it to its environment. We present the design and prototyping of the complete system and validate its performance through a series of real-robot flight tests. Results demonstrate that anchoring significantly improves stability for interaction tasks, reducing positional drift RMSE by over 95% compared to a free-flight baseline, even under windy conditions. The anchored system can withstand longitudinal reaction forces up to 75N while maintaining a stable pose. Furthermore, across a range of target geometries and orientations, the system demonstrated consistent stability with a positional drift RMSE that never exceeded 3mm. These results establish the viability of our approach for complex physical interaction tasks, such as sampling tree health by drilling or sensor installation in hard-to-reach locations. Watch our UAV at: https://youtu.be/HDQ8S4ZW3Ls
Multi-link aerial robots can actively deform their articulated structures during flight, giving them strong potential for aerial manipulation. However, they still face substantial challenges in contact-rich aerial manipulation tasks such as surface sliding, which requires both disturbance robustness and compliance to uncertain surface geometry. Force-control strategies such as impedance and admittance control are commonly employed to address these requirements. Although impedance control can provide disturbance-resistant interaction and admittance control can offer compliant adaptation, their opposite force--motion causalities prevent their simultaneous implementation when applied through the same actuation source, such as the rotor thrusts used by conventional aerial robots. To overcome this limitation, we propose a hybrid impedance--admittance control strategy for a multi-link aerial robot. The articulated morphology enables a functional separation of force and motion regulation across joint and rotor actuation sources. In this framework, admittance behavior is generated through joint angle regulation to enhance adaptive interaction, while impedance behavior is achieved by modulating rotor thrust to regulate the sliding motion. This structural coordination allows the robot to leverage the complementary strengths of both control paradigms. As a result, the multi-link aerial robot achieves resilient and adaptive surface sliding. Experimental results demonstrate robust and compliant sliding performance on unknown surfaces.
Fusing tactile signals has proven effective for contact-rich manipulation, enabling robots to perceive contact states and adapt to rapidly changing physical interactions. Yet effectively integrating tactile feedback into dexterous manipulation remains underexplored. In this work, we introduce ReTouch, a vision-language-action model (VLA) that supports contact-rich dexterous manipulation through tactile predictions continually refined online using execution-time feedback. ReTouch builds on two main innovations for tactile representation and closed-loop action generation. First, its Tactile-Patch Encoder represents tactile observations as structured tactile patch features that preserve finger identity and local contact structure, providing contact cues for fine-grained dexterous control. Second, its high-frequency action module jointly predicts future tactile states and action chunks and refines both using incoming tactile feedback during execution. This closed-loop refinement keeps tactile predictions aligned with evolving physical interactions, enabling responsive action correction and improving robustness to contact changes and execution errors. We further introduce XHT-Dataset, comprising 900 real-world demonstrations across seven contact-rich tasks collected on an XHand--UR7e platform, and evaluate ReTouch through closed-loop real-robot experiments. ReTouch surpasses the strongest baseline by 18.4 and 23.8 percentage points in average success rate under standard and challenging conditions, respectively, demonstrating its effectiveness and robustness.
Vision-Language-Action (VLA) models have shown strong generalization in robotic manipulation, yet complex contact-rich tasks often benefit from multi-camera observations that jointly capture the end effector, objects, and targets under occlusion. Existing multi-camera VLAs usually concatenate view tokens, leaving action representations weak in metric depth and inconsistent across cameras. We introduce Multi-View Unified Camera Fields (MVUCF), a training-only framework that forms a shared action-facing latent field across views. A coordinate-query depth objective makes metric depth recoverable, while a preprocessing-aware correspondence objective aligns tokens observing the same physical point from different cameras. Both directly shape the hidden states consumed by the action module. After geometry injection, depth, camera calibration, and auxiliary heads are removed, so deployment uses the original RGB-only graph with no extra inference FLOPs. Held-out probes confirm stronger depth recovery and cross-view matching. Under matched GR00T-N1.6 settings, MVUCF reaches 98.9% on LIBERO, improves LIBERO-Plus by 22.4 points, and raises success by 23.3 points across six RoboTwin tasks spanning three action families: touch, move-and-place, and contact interaction. Real-world humanoid experiments further provide evidence of its practical effectiveness under RGB-only deployment.
Humanoid teleoperation for demonstration collection requires coordinated whole-body motion, continuous dexterous hand control, and viewpoint control. Existing systems either simplify hand commands or depend on dedicated wearable sensors for fine-grained hand motion. We introduce Teleopit, a full-embodiment teleoperation system that maps body, hand, and head signals from VR to a humanoid body, configurable dexterous hands, and a 2-DoF active vision module. A history encoder and failure-aware rewind sampling improve the motion tracker on both motion-capture and live VR references. An optimization-based hand retargeter combines normalized finger directions, fingertip closure, and thumb-frame alignment to map human hand motion to different dexterous hands without tuning hand-specific objective or solver hyperparameters. Component experiments evaluate tracking success rate and retargeting behavior, while real-robot teleoperation demonstrates coordinated locomotion, manipulation, and viewpoint control. ACT and GR00T N1.7 policies trained on 96 successful demonstrations collected with Teleopit achieve task success rates of 90.0% and 95.0%, respectively, when deployed on the humanoid. The project page is available at https://botrunner64.github.io/teleopit-page.
Robot learning is splitting into two bets: policies that bake competence into frozen weights (vision-language-action, or VLA, models), and agents that write and refine their own executable skills as code. This survey organises the field around that axis of weights versus skills. Its central analytical contribution is a deep-dive that arranges code-as-policy methods by their degree of self-improvement, from zero-shot program synthesis, through closed-loop self-repair and persistent skill memory, to the sparsely populated cell in which execution feedback, skill memory, and evolutionary search combine into one open-ended loop; only a few very recent systems (for example ASPIRE, ENPIRE, and RoboClaw) occupy that cell. We map the complementary "skills" pole, from unsupervised reinforcement-learning skill discovery to large-language-model skill libraries, and show that the word "skill" is used in at least five distinct senses, of which only the code sense self-improves without gradient updates. We then connect the taxonomy to the emerging skill economy: commercial robot-skill marketplaces now distribute one-tap skills across robots but ship only static playback, which surfaces open problems of adaptation, cross-embodiment portability, provenance, safety verification, composition, and standardisation. This is a deliberately focused survey. Rather than cataloguing the field exhaustively, it examines 77 representative systems across six technique families through one taxonomy and a set of contrast tables, and it supplies operational definitions of the self-improvement mechanisms together with a statement of what each family cannot do.
RADAR Perception for Dynamic Obstacle Avoidance onboard small-scale Quadrotor UAVs
基于雷达的动态障碍物规避系统在小型四旋翼无人机上的应用
Mandaokar, Dnyandeep, Rinner, Bernhard
Abstract
Fast dynamic obstacle avoidance (DOA) on uncrewed aerial vehicles (UAVs) demands not only low-latency control and actuation but also reliable perception with sufficient sensing range for accurate obstacle detection and speed estimation. This letter presents, to the best of our knowledge, the first mmWave RADAR-based perception-and-control system for fast onboard DOA. We derive and analyze latency and spatial bounds that relate sensing range, relative speed, and control delay, yielding sufficient conditions for successful avoidance. Our system adopts a lightweight tracker based on interacting multiple models and a controller based on control-barrier functions that directly outputs evasive accelerations. It achieves position errors of less than 0.15 m, 0.93 m, and 0.87 m in x, y, and z directions for 300 experiments with three different object sizes and varying visibility (light and dark), and a similar spread for 90 experiments in smoke. An onboard implementation on a Raspberry Pi 4B demonstrates real-time feasibility with an end-to-end sensing-to-command latency of approximately 14 ms. Code and the full dataset of 390 throws are available (https://tinyurl.com/radardoagit).
World Action Models in Real Time: An Empirical Study of Smooth Execution via Asynchronous Deployment
实时世界动作模型：通过异步部署实现平滑执行的实证研究
Motubrain Team
Abstract
World Action Models generate fixed-horizon action chunks through iterative denoising, creating substantial inference latency that can cause pauses, stale actions, and discontinuities during robotic execution. We present an empirical study of asynchronous deployment strategies that overlap model inference with action execution to enable responsive and smooth control. We compare six strategies, including synchronous execution, pure asynchronous switching, post-hoc action blending, denoising-time blending, inference-time velocity guidance, and prefix-conditioned generation, on a 10 Hz bimanual robot. Evaluation combines offline trajectory analysis with online experiments across dynamic manipulation, precision-critical placement, and long-horizon tasks. Our results identify accurate temporal alignment between observations, predictions, and executed commands as a fundamental requirement. Alignment errors produce persistent chunk-boundary discontinuities that cannot be corrected through blending alone. With proper alignment, direct action weighting provides a simple and smooth baseline but sacrifices accuracy in precision-critical tasks. Inference-time velocity guidance fails to reliably constrain committed actions on our platform. In contrast, prefix-conditioned generation achieves the best overall balance between task performance, execution speed, and trajectory smoothness by learning consistent action continuations during training. These findings clarify the practical trade-offs among asynchronous deployment strategies and provide guidance for deploying high-latency World Action Models in real-time robotic systems.
FRA-NBV: A Fast and Reflectivity-Aware Next-Best-View Strategy
FRA-NBV：一种快速且考虑反射的下一最佳视角策略
Preziosa, G. F., Setti, E., Faroni, M., Zanchettin, A. M., Rocco, P.
Abstract
Autonomous 3D reconstruction with depth sensors is strongly affected by reflective surfaces, which cause missing or unreliable measurements and reduce the effectiveness of conventional Next-Best-View (NBV) strategies. This limitation is particularly critical in industrial applications involving reflective components and low-cost, low-resolution depth sensing, where robustness to sensing failures is essential. This paper proposes a Fast Reflectivity-Aware Next-Best-View (FRA-NBV) strategy that explicitly addresses reflection-induced depth loss without relying on prior object models or assumptions on material reflectance, making it suitable for a wide range of industrial configurations. Reflective regions are identified from the spatial distribution of missing depth measurements and localized in three-dimensional space using an online ellipsoid-based representation of the object estimate. A recovery strategy then selects additional poses that modify the sensor's angle of incidence to improve the likelihood of reconstructing the affected regions. Experiments on objects with different geometric and reflective complexity demonstrate that the approach significantly improves reconstruction coverage under realistic industrial conditions.
Existing indoor layout generators produce globally plausible layouts yet may retain local violations such as collisions, out-of-bounds placements, obstructed openings, and blocked circulation. Most prior work focuses on full-scene synthesis or scene-level optimization, with limited support for identifying responsible objects and locally repairing affected regions. We present Roomer, a reflective repair framework that casts these violations as sparse, object-grounded repair problems. Roomer encodes layouts as ``RoState'' and uses ``RoReview'' to bind measured violations to implicated objects. A geometry-conditioned vision-language model planner proposes a structured local edit, while a deterministic solver validates it and generates a finite set of candidate edits when needed. Each candidate is committed only if full-scene verification confirms that it resolves the target violation without new hard violations or broken protected constraints. We train the planner on Roomer-CC, a controlled-corruption dataset that pairs faulty layouts with object-grounded violation evidence and known-feasible inverse StatePatches. Since existing benchmarks rarely assess whether physically valid layouts are usable, we introduce Roomer-Eval to assess distributional quality, physical validity, and practical usability. Experiments show that Roomer repairs residual violations while preserving valid regions, improves physical validity and usability, and transfers across external generators.
Cella, C., Ristic, M., Faroni, M., Zanchettin, A. M., Rocco, P.
Abstract
This work presents a human-centered collaborative framework that integrates Preference-Based Optimization (PBO) and Dynamic Movement Primitives (DMPs) to optimize robot-assisted tasks such as painting. The system allows the operator to perform the process while the robot adapts its behavior in real-time, dynamically adjusting the orientation of the piece in order to match the orientation of the operator's hand. The PBO framework leverages the GLISp algorithm to iteratively refine control parameters such as execution time, robot responsiveness, and rotation amplification through human feedback. Moreover, DMPs have been modified to enhance the reactive behavior of the robot and its adaptability to ergonomic requirements. The method was validated with a heterogeneous group of participants executing \rev{painting tasks}. The results show that our strategy effectively reduces operator effort while optimizing process outcomes.
Chinese Translation
本研究提出了一种以人为中心的协作框架，该框架整合了基于偏好的优化（Preference-Based Optimization, PBO）和动态运动原语（Dynamic Movement Primitives, DMPs），以优化机器人辅助的任务，例如绘画。该系统允许操作员在绘画过程中进行操作，同时机器人实时适应其行为，动态调整作品的方向，以匹配操作员手部的方向。PBO框架利用GLISp算法，通过人类反馈迭代优化控制参数，如执行时间、机器人响应能力和旋转放大。此外，DMP经过修改，以增强机器人的反应行为及其对人体工程学要求的适应性。该方法在一组异质参与者执行绘画任务时得到了验证。结果表明，我们的策略有效降低了操作员的工作量，同时优化了过程结果。
MANGO-Grasp: Mahalanobis Fields over Geometry-Oriented 3D Gaussians for Cross-Embodiment Dexterous Grasping
MANGO-Grasp：基于几何导向的3D高斯分布的马哈拉诺比斯场用于跨体现灵巧抓取
Zhang, Heng, Ma, Kevin Yuchen, Shou, Mike Zheng, Lin, Weisi, Wu, Yan
Abstract
Cross-embodiment dexterous grasping aims to synthesize stable grasps across heterogeneous multi-fingered hands with little or no embodiment-specific tuning. Existing interaction-centric methods achieve promising results, but their object representations often underrepresent local surface geometry, while their robot descriptors do not explicitly encode both robot morphology and kinematics. We propose MANGO-Grasp, an anisotropic interaction framework that represents objects as geometry-oriented 3D Gaussian primitives and robot hands as surface keypoints encoded into morpho-kinematic descriptors. The object primitives are adaptively allocated by geometric complexity and shaped as surface-aligned plates with outward normals, encoding local geometry. Mahalanobis fields over keypoint--primitive pairs serve as interaction prediction targets during training and as optimization guidance for grasp realization at inference. These fields rise sharply for displacement along the surface normal but only gently within the tangent plane, matching the directional structure of contact. Grasps are realized with one shared optimization formulation and hyperparameter setting across all embodiments. On the CMAP and MultiGripperGrasp benchmarks, MANGO-Grasp outperforms the strongest seen-hand baseline by up to 8.24 percentage points in simulation. It also transfers zero-shot to the unseen SharpaWave hand, improving over the strongest zero-shot baseline by up to 16.57 percentage points, and achieves 86% success in real-world experiments. The code and additional materials will be made available upon publication at https://connor-zh.github.io/MANGO-Grasp/.
Open-DiffLoco: Open-Source Differentiable Learning for Deployable Blind Quadruped Locomotion
Open-DiffLoco：可部署盲四足运动的开源可微学习
Opat, Martin
Abstract
Developing deployable locomotion policies through conventional reinforcement learning often requires complex reward engineering and expensive training times. While differentiable simulation offers a highly efficient alternative, open-source tools capable of end-to-end transfer of these policies to physical hardware remain limited. This paper introduces Open-DiffLoco, an open-source framework for training deployable blind quadruped locomotion policies with differentiable simulation. The framework implements the Short-Horizon Actor-Critic (SHAC) algorithm in MuJoCo XLA (MJX) and trains a proprioceptive policy that transfers to real-world hardware. The deployed policy removes privileged actor observations, including base linear velocity, and does not rely on reference trajectories. It also uses a substantially simplified reward function, enabling the robot to discover walking patterns without the complex auxiliary rewards typically used in conventional reinforcement learning pipelines. When deployed on physical hardware (a Unitree Go2 quadruped), the trained policy tracks omnidirectional velocity commands with root-mean-square error below 0.2 m/s, reaches speeds above 1 m/s, and remains robust to uneven terrain and external physical disturbances, such as lateral pushes. Across the reported configurations, training uses under 6 GB of VRAM on a single NVIDIA GeForce RTX 5080 GPU and completes in approximately 20-60 minutes. As an algorithmic extension to SHAC, we propose Jacobian-Augmented Value Estimation (JAVE), which supervises the critic Jacobians to improve early first-order policy-gradient training. To our knowledge, Open-DiffLoco is the first open-source framework for training deployable locomotion policies using differentiable simulation. Deployment videos and source code are available at: https://diffloco.martin-opat.com/
TANGO-VIO: Triangulation-Aware Navigation with Guaranteed Feature-Observability for Visual-Inertial Odometry
TANGO-VIO：具有保证特征可观测性的三角测量感知导航用于视觉惯性里程计
Altunkaya, Ege C., Şanlan, Abdülbaki, Koyuncu, Emre, Özkol, İbrahim
Abstract
In vision-aided navigation and visual-inertial odometry, the quality of triangulated three-dimensional feature positions is a fundamental prerequisite for state estimation accuracy. Triangulation becomes ill-conditioned or even impossible when a camera undergoes pure rotation without translation, or when the observed bearing vectors provide insufficient parallax. Even though visual-inertial odometry has been extensively studied, the active maintenance of feature-observability during navigation has not been sufficiently addressed in the literature. To address this gap, this study presents TANGO-VIO, a triangulation-aware navigation framework that embeds a log-determinant metric of the feature-wise stacked-bearing matrix into a control barrier function. In this proposed method, the observability guarantee is established in the feature-geometric sense by enforcing a lower bound on the aggregate triangulation-information metric through a nominal-direction-weighted minimum-deviation velocity correction. The proposed architecture is evaluated through software-inthe- loop simulations and real flight experiments. The results show improved triangulation conditioning under low-parallax motion, while the flight response closely reproduces the corresponding simulation behavior and confirms the practical realizability of the proposed safety filter. Supplementary materials are available on the project webpage.
Whole-body tactile sensing is a prerequisite for humanoids that operate in contact-rich human environments, but conventional taxel arrays scale poorly with surface area, wiring complexity, and robot-specific curvature. We present a conformal electrical impedance tomography tactile skin fabricated through a geometry-adaptable additive-manufacturing workflow. A flexible conductive TPU layer forms a continuous sensing domain, while contact-induced coupling with conductive patches produces boundary voltage changes that are reconstructed using a one-step Gauss-Newton EIT solver. We first characterize the electromechanical design space of the layered structure and show that low-resistance contact-enhancement patches and a porous conductive TPU sensing layer improve sensitivity while preserving printability. We then validate contact localization on a planar prototype, a curved U-shaped prototype, and a qualitative iCub-face-shaped geometry. The curved sensor achieves a mean localization error of 6 mm over 18 contact positions without supervised post-processing. These results suggest that additively manufactured tomographic skins can reduce the morphology-specific redesign burden for humanoid tactile coverage and provide a practical route toward large-area contact sensing for human-centered deployment.
Dynamic UAV-based search operations using probabilistic diffusion modeling of Man Overboard incident victims
基于动态无人机的搜索操作：使用概率扩散模型预测落水事件受害者
Angelis, Dimosthenis, Boukas, Evangelos
Abstract
More than 70% of the people that fell overboard cruise ships in the period 2010-2019 lost their lives. This paper presents a strategy for reliably predicting the area a person may be in after a man overboard incident, and describes in detail the search methods to find them utilizing UAV technology. The search area prediction method employs an Extended Kalman Filter that capitalizes on the information from the Leeway model to track the missing person in the sea by taking into account the uncertainty of the movement of the person and the weather conditions in the area. Then, a UAV uses this information to search for the person. Five different methods for searching in this dynamic area are presented and evaluated - the Zigzag, the Boustrophedon, the Spiral, the Probability Informed Search and the Improved Probability Informed Search (IPIS) methods. The IPIS method provides success rate of over 80% on average for finding a person, even if the UAV initiates the search mission 20 minutes after the man overboard incident and even assuming a detection method with a success rate of 30%. All code for the simulation environment and the evaluation of the methods is available on our GitHub page at https://github.com/diangeli/pdms-man-overboard.
Visual representations of VLA models remain unreliable for spatially precise robotic manipulation. We uncover that vision encoders in VLAs also exhibit attention artifacts previously documented in generic Vision Transformers, and further show that, in embodied policies, these artifacts are closely associated with spatial perception capabilities acquired during post-training. As the encoder learns task-relevant information such as object location, depth ordering, and local geometry, limited global-token capacity causes part of this information to spill into low-information patch tokens. We introduce AtVLA, a framework that inserts learnable register tokens into the visual encoder. Trained end-to-end using only embodied data and the original action objective, these registers emerge as dedicated carriers of embodied spatial information, while the remaining patch tokens recover clean and spatially faithful attention distributions crucial for precise target localization and fine-grained contact. Clean attention restores reliable localization, but cannot recover geometric details lost in low-resolution observations. AtVLA therefore couples attention rectification with uncertainty-gated local refinement. The action expert samples multiple action chunks and estimates uncertainty from their disagreement; only for uncertain predictions, action-conditioned attention rollout identifies the task-relevant region, which is cropped, re-encoded at high resolution, and appended to the cached prefix for refined action generation. Across LIBERO, SimplerEnv, and a challenging single-view real-world benchmark, AtVLA improves the average LIBERO success rate from 94.2% to 98.4% and real-world success from 46.5% to 69.0%. The cropping is triggered on approximately 30% of replanning steps, resulting in only 1.4-1.6x the total computation of the base model under the representative deployment setting.
Mobile manipulation is a key capability for embodied intelligence, enabling robots to accomplish complex multi-stage tasks in open-world environments. However, mobile manipulation poses two key challenges for vision-language-action (VLA) policies: At the data level, the efficient collection of high-quality whole-body demonstrations demands the coordinated control of both the mobile base and the robotic arms; at the model level, existing VLA models predominantly rely on local camera observations, whose limited field of view hinders global spatial understanding. To address these challenges, we develop a whole-body teleoperation system and a panoramic-aware VLA policy. The system enables coordinated control of a wheeled bimanual robot through a single VR interface and supports the acquisition of a real-world mobile manipulation dataset comprising 5.5 hours of multimodal demonstrations. Building upon this dataset, we propose PanoVLA, a panorama-aware vision-language-action policy for mobile bimanual manipulation. Built upon a Mixture-of-Transformers architecture, PanoVLA introduces global spatial context through dedicated panorama encoding and fusion modules, enabling effective integration of panoramic observations with language instructions and robot states for action generation. Evaluation on four real-world mobile manipulation tasks demonstrates that PanoVLA achieves an average stage completion rate of 91.3\% and an end-to-end success rate of 73.4\%, substantially outperforming local-view baselines. These results demonstrate that incorporating panoramic spatial context improves spatial understanding and closed-loop manipulation performance in mobile robots.
Agriculture 4.0 robotic systems improve field efficiency yet remain too capital-intensive for the fragmented smallholdings that dominate global agriculture. Meanwhile, a growing number of retired low-speed electric-vehicle (LSEV) powertrains retain functional electromechanical value but are destructively recycled. This paper presents TS-MAMP (Telescopic-Sleeve Modular Agricultural Mobile Platform), a remanufactured robot built under 3R (reduce, reuse, recycle) circular-economy principles. Retired 48 V brushless-DC (BLDC) hub motors are paired via back-EMF matching, and lead-acid battery modules screened at 60%-80% state of health are actively balanced within a 100 mV inter-module voltage deviation. Together, these reused components reduce the powertrain-and-chassis BOM cost by approximately 60%, to below USD 450 (perception and weeding modules excluded). The truss chassis provides >=200 kg static load, continuously adjustable track width from 1200 mm to 2000 mm, and <=5-minute module changeover. An NMS-free (non-maximum-suppression-free) YOLOv10n detector with consistent dual-assignment training and negative-sample learning achieves 80.87% mean average precision (mAP)@0.5 (58.41% [email protected]:0.95) on the Wanxi Crop-Weed dataset, and is deployed via FP16 TensorRT on a Jetson Nano, confirming on-device inference feasibility. TS-MAMP demonstrates that retired EV components, under modest screening, can be re-engineered into affordable, AI-enabled agricultural robots--opening a remanufacturing pathway for the smallholder fields that commercial automation leaves unserved.
Existing active reconstruction systems with Gaussian-splatting maps select observations greedily, optimizing a single next-best-view (NBV) at each step and connecting the chosen views by short-horizon path planning. This greedy decoupling disregards the global structure of scene information, producing inefficient trajectories that waste sensing capacity in transit between selected views. In this work, we study active reconstruction as an ergodic coverage problem: the time-averaged spatial statistics of the sensor trajectory should match a target information distribution induced by the current map. Our approach derives this target distribution online from uncertainty and visibility, and calculates ergodic trajectories via a kernel-ergodic horizon planner with gradient flow and footprint depletion, closing the loop between mapping and trajectory optimization. We thoroughly evaluate TRACE on the Replica dataset against the Next-Best-View (NBV) baselines, improving PSNR by 1.5 dB. Code: https://github.com/spikelab-jhu/trace-active-reconstruction.
Residual-Based Adaptive Kalman Filtering for Legged Robot State Estimation
基于残差的自适应卡尔曼滤波用于腿式机器人状态估计
Popescu, Mihaela, Mronga, Dennis, Kumar, Shivesh, Kirchner, Frank
Abstract
State estimation is a key component in model-based control of walking robots and, more broadly, applicable wherever hidden variables must be inferred. The Kalman filter is widely used to estimate floating-base position and velocity by fusing multiple sensing modalities. However, tuning noise parameters is challenging and typically requires expert knowledge. Moreover, fixed noise parameters are unsuitable for varying gaits and environments. We propose an online adaptation strategy for the process noise covariance matrix Q and the measurement noise covariance matrix R. Specifically, we introduce a filter residual and innovation-based covariance adaptation method for legged robot state estimation and evaluate it against a baseline approach relying on IMU and foot force measurements. The proposed adaptation is implemented within an Invariant Extended Kalman Filter (InEKF) fusing IMU and leg kinematics. Experiments on indoor and outdoor datasets with a Unitree Go2 quadruped show that adapting R is sufficient and improves accuracy by 25% for the trotting gait compared to the fixed-tuned InEKF. Finally, the proposed residual-based adaptation achieves comparable performance to the foot force approach, without requiring foot force measurements or additional parameter tuning.
TravKAN: Fast and Interpretable Nonlinear Traversability Analysis with Kolmogorov-Arnold Networks
TravKAN：基于Kolmogorov-Arnold网络的快速且可解释的非线性可通行性分析
Fusaro, Daniel, Mosco, Simone, Li, Wanmeng, Pretto, Alberto
Abstract
Traversability analysis is a fundamental capability for autonomous mobile robots operating in unstructured environments. While modern machine learning approaches such as deep neural networks and gradient-boosted trees achieve strong predictive performance, they lack interpretability and provide limited insight into the underlying terrain-robot interaction dynamics. In this paper, we propose TravKAN, a Kolmogorov-Arnold Network-based framework for fast, scalable, and interpretable traversability estimation. TravKAN represents multivariate decision functions through compositions of learnable univariate functions, enabling compact architectures and symbolic extraction of analytic expressions after training. In addition, we introduce a novel set of handcrafted features derived from the reflectivity channel of LiDAR sensors. To the best of our knowledge, reflectivity has not been systematically exploited for handcrafted traversability descriptors, despite its potential to capture material and surface properties complementary to geometric cues. We evaluate TravKAN on public, real-world urban and off-road datasets and compare it against strong baselines. TravKAN achieves strong performance across all metrics, outperforming conventional deep models and approaching the performance of XGBoost. TravKAN-Lite, i.e., TravKAN's symbolic representation, reveals meaningful nonlinear feature interactions and provides a compact, deployment-friendly, and fast analytic model. Ablation studies further show the robustness of our method to architectural variations and quantify the contribution of the proposed reflectivity-based features. These properties make TravKAN attractive for robotic systems requiring transparency, real-time computational efficiency, and interpretability in safety-critical decision-making.
Humans perform long-horizon manipulation by retaining knowledge of what earlier actions have established while continuously adapting the motion underway. By contrast, action-chunked vision-language-action (VLA) policies repeatedly replan from the current input at each query. Existing methods preserve either long-term task evidence through memory or short-term motion through action reuse and ensembling, leaving the cross-query handoff incomplete. We introduce ChainVLA, a 1.2B-parameter VLA policy that chains successive queries through a joint and revisable execution state. Progress Context combines a recurrent Working State with sparse event memory to carry observation-derived task progress, while Motion Tail feeds the preceding prediction's unexecuted continuation into state construction and action generation. Together, the two components condition a decoder that regenerates each action horizon under the latest observation, allowing the carried state to guide the next prediction without fixing it. ChainVLA reaches 62.8% average success on RMBench and 98.8% across four LIBERO suites, while removing Motion Tail or Progress Context reduces RMBench success to 11.2% and 3.0%, respectively. These asymmetric ablations are consistent with motion continuity helping preserve the observation stream from which task progress is inferred.
Humanoid motion trackers perform reliably within learned tracking distributions, but falls can move the robot into low-height, contact-rich states from which an advancing command is temporarily unreachable. Tracking-only policies may chase infeasible references, producing rapid, large-amplitude limb corrections that increase risk to the robot and its surroundings. We present StableMimic, a unified tracker trained beyond the nominal tracking distribution. Perturbed resets around multiple human get-up references expose prone, supine, off-balance, and intermediate ground-contact states, shaping structured recovery that returns the robot to the trackable region. Because tracking and recovery occupy markedly different state--action distributions, StableMimic uses dedicated experts for each regime and a proprioceptive gate that continuously blends their actions. A hidden successor-state objective teaches human-reference-shaped recovery without exposing reference identity or phase to the deployed Actor; deployment requires no get-up reference, recovery command, trajectory retrieval, or external policy switch. On the complete retargeted LAFAN1 dance subset, StableMimic achieves the lowest errors on all four tracking metrics among five methods. Across 100 matched push-to-fall trials per method, it recovers in 100/100 and attains the lowest values on six of seven post-fall motion and load measures, supporting improved interaction safety under this protocol. Real Unitree G1 dance and standing-reference deployments qualitatively demonstrate bounded limb motion, autonomous recovery, and command resumption.
Environmental resilience via morphological diversity within machines
通过机器内部形态多样性实现环境韧性
Hein, Alice, Bongard, Josh
Abstract
Organisms contain diverse, sensorimotor parts across size scales and rapidly adapt to new environments, while machines contain only inert materials at smaller scales and struggle with surprise. We hypothesize that this agents-within-agents quality of organisms may aid their resilience: increasing experiences with internal physical adversity may pre-train organisms and machines to handle external adversity, such as encounters with new environments. Not only has this hypothesis not yet been articulated, mechanisms enabling this phenomenon have yet to be proposed. Here we show a mechanism by which this can occur: we found that physical connectors, in learning to restore behavior to previously independent, morphologically diverse agents they disrupted by tethering them together, trigger and tame sufficiently diverse disruptions that later encounters with new environments trigger disruptions that fall within this manageable range, enabling the collective to continue behaving properly without any additional learning or adaptation. Further, we found that building collectives from more agents, or more diverse agents, further increases the collective's resilience to new environments. This suggests that not just taming but intentionally creating internal physical adversity may indeed prepare organisms for external adversity, and could do so for machines, if they were built from smaller machines.
Human-Centered Reflections on Care Robots: A Comparative Study of Caregiver Perspectives
以人为本的护理机器人反思：护理人员视角的比较研究
Londoño, Laura, Baumann, Klaus, Valada, Abhinav, Langer, Markus
Abstract
Care robots are increasingly being introduced into healthcare settings, raising important questions about their acceptance and ethical implementation. To better understand these challenges, this study investigates caregivers' perceptions of four categories of care robots: delivering supplies, helping patients into bed, monitoring vital signs, and assisting with mobility. We conducted a mixed-methods study employing a mixed-factorial design in which 298 caregivers from the United States, Mexico, and Chile evaluated all four robot categories. Quantitative measures integrated constructs from the Unified Theory of Acceptance and Use of Technology, the Cognitive-Affective-Normative model, and overall acceptance ratings. Qualitative data were collected through open-ended questions and analyzed using a literature-informed ethical framework. The results indicate that participants across countries generally evaluated care robots positively, particularly for logistical and physically demanding tasks rather than those requiring intensive interpersonal interaction. The qualitative findings provide further insight into stakeholders' views of the ethical implications of care robot use. Participants emphasized potential benefits such as reduced workload, lower risk, and greater patient autonomy, while also expressing concerns about dependability, the need for human oversight, and potential job displacement. Although many ethical concerns were shared across countries, participants differed in how they interpreted and prioritized them. These findings advance a context-sensitive and socially informed understanding of responsible design and implementation of care robots.
Sim-to-real policies are designed under nominal dynamics, but target-system trials may yield only a few isolated one-step transitions. We study pre-execution certification of a fixed control sequence, such as an action chunk produced by a learned policy. If the sequence reaches an unobserved state-input region, the observations remain consistent with target systems whose trajectories separate along it by an arbitrarily large amount. Any deterministic certifier sound for all of them must then decline to certify or return a reachable tube with arbitrarily large projected width. For bounded smooth classes of the target-nominal model error, we derive a finite plan-dependent projected-width lower bound. These results expose a trilemma among uniform trajectory containment, finite projected width, and unrestricted model-error behavior beyond the observations. ForeReach requires a supplied componentwise Lipschitz bound on the model error. Observed transition pairs can refute this declaration but cannot establish it outside the observed locations. Conditional on a valid declaration, our method constructs a set-membership envelope for the model error, propagates a zonotopic reachable tube, and certifies only when propagation remains within the certification domain and every projected tube slice avoids the unsafe set. In two benchmark systems, calibration baselines may remain narrow after losing trajectory containment outside data support, whereas our method declines to certify unsupported sequences and recovers certification when relevant target data and sufficient obstacle clearance are available.
Grounded Semantic Re-Binding for Robust Instruction Generalization in Vision-Language-Action Models
基于语义重绑定的稳健指令泛化方法在视觉-语言-动作模型中的应用
Yin, Zhaokai, Zhang, Zhipeng
Abstract
Vision-Language-Action (VLA) models excel in robotic manipulation but suffer catastrophic performance drops when canonical instructions are simply paraphrased. Although this brittleness is typically addressed through costly data scaling, our probing reveals that the root cause is architectural rather than a lack of semantic understanding. Specifically, we demonstrate that current VLAs successfully retain the correct task identity internally. The failure actually stems from the joint encoding of dynamic visual observations and text, which introduces systematic feature shifts. Because the downstream action policy is highly vulnerable to these variations, it fails to translate the preserved semantics into correct control commands. To resolve this structural bottleneck, we propose Grounded Semantic Re-binding (GSR), an elegant intervention that bypasses unstable joint routing by explicitly fusing independently extracted task semantics with native visual features to train a completely re-initialized action expert from scratch. This targeted intervention dramatically restores paraphrastic invariance using only canonical demonstrations. On the LIBERO-Para benchmark, GSR improves success rates by up to 44.6 percent. It enables lightweight models to rival massively scaled baselines and pushes state-of-the-art models to a new record PRIDE score of 70.4, outperforming the recently introduced large-scale pretrained model Xiaomi-Robotics-0 in instruction generation capabilities. Building on these insights, we also introduce ParaVLA, a natively decoupled 0.33B-parameter model exhibiting near-perfect robustness to instruction rewording. Ultimately, our work proves that robust semantic grounding can be achieved through elegant structural design, bypassing the inefficient brute-force data scaling paradigm.
Reachability analysis for visuomotor policies is difficult because large visual encoders make end-to-end set propagation computationally expensive and excessively conservative. We therefore freeze the visual encoder and confine set propagation to a low-dimensional interface between it and the downstream policy, with the interface set calibrated from held-out camera-pose perturbations. Propagating this set through the policy with zonotopes yields a terminal output-enclosure width that set-based training optimizes directly. During evaluation, camera-pose perturbations are sampled from the prescribed distribution, and rollout-level split conformal calibration converts the resulting action-deviation scores into a probabilistic reachable-action radius with finite-sample coverage. In controlled manipulation experiments, set-based training reduces this radius while preserving closed-loop task capability, and matched behavior-only, observational-consistency, and pointwise-adversarial controls all leave a larger radius.
Action chunking---predicting and executing multiple actions instead of a single action---has proven to be a critical component for learning effective robotic control policies. However, our precise understanding of why action chunking improves performance has remained limited. In this work we seek to close this gap. Through rigorous experimental evaluations in both simulated and real-world settings, we show that existing hypotheses for the success of action chunking---temporal consistency, horizon reduction, and representation learning---fail to explain the success of action chunking. Instead, we find that action chunking benefits from greater non-Markovian expressivity and reduced compounding error compared to Markovian policies, but, in many settings of interest, these effects can be fully captured by delayed policies, which at each step predict a single action based on the observation $k$ steps in the past. We then show that there exists an additional benefit of action chunking that we refer to as implicit ensembling. In particular, by learning a diversity of temporal relationships (that is, $a_t | o_t, a_t | o_{t-1}, \ldots$), action-chunked policies exhibit behavior matching that of a model ensemble, increasing their robustness and generalization ability over policies that only learn a single temporal relationship. Building on these insights, we show that in simulated and real-world robotic control settings, we can match the performance of action chunking without action chunking---by deploying an action chunking policy as an ensemble of policies with randomized delays. Furthermore, we propose a policy class that amplifies the benefits of action chunking by explicitly instantiating an ensemble, and which we show significantly improves over the performance of action chunking in many domains.
Quadruped robots are a promising platform for search and rescue missions because they can navigate cluttered indoor environments that may be restrictive for wheeled systems. However, in unknown rescue scenarios, autonomous exploration must balance map expansion with the likelihood of finding victims, which is not explicitly addressed by clas- sical frontier selection strategies. This paper presents a situation aware frontier prioritization method for single robot quadruped search and rescue. The proposed approach preserves the frontier exploration framework, but extends frontier ranking with information gain, observation deficit, rescue relevance, terrain penalty, and travel cost. The method is eval- uated in Gazebo simulation with a quadruped robot in two indoor rescue scenarios with different levels of difficulty. The first scenario is used as a sanity check, while the second introduces stronger clutter and frontier ambiguity. Experimental results show that all methods perform reliably in a simple scenario, whereas in a complex scenario is different. In that setting, the proposed method achieves the highest completion rate and the highest victim recovery among the evaluated approaches. These results indicate that situation aware frontier prioritization is beneficial when frontier choice becomes nontrivial and rescue utility must be balanced against generic exploration objectives.
World Action Models (WAMs) augment robot policies with action-conditioned predicted futures, but a plausible future alone does not justify changing the action that a bimanual policy would execute. We present CoWAM, a selective intervention layer that expresses synchronization, role compatibility, and collision convergence as coordination contracts. Each contract combines typed admissibility checks with event-conditioned verification and calibrated intervention gates. CoWAM preserves the nominal action unless an alternative satisfies every active obligation and provides a clear, low-risk improvement; when the nominal action is also inadmissible, it invokes a predefined abstention fallback. To separate selector quality from proposal quality, all methods operate on identical candidate pools and commit their decisions before shared oracle labeling. Across eight simulated bimanual tasks, CoWAM improves coordination-valid selection by 16.7 percentage points over the contract-only variant and raises closed-loop success by 9.6 percentage points over the strongest selective baseline, while keeping harmful interventions below 1%. Together, these results establish coordination contracts as an effective interface for conservative policy intervention with predicted world-action evidence across coordination-rich bimanual tasks.
Learning generalizable robot manipulation policies requires large-scale and diverse demonstration data. Egocentric human manipulation videos offer rich scene and task diversity, and prior work has shown that retargeting and rendering such videos into robot-format data can yield effective per-task policies at small scale. However, whether this approach can provide pretraining benefits for vision-language-action models at scale remains unexplored. We present \textbf{Ego2Robot}, a scalable pipeline that converts egocentric human manipulation videos into robot training data through action retargeting, robot-arm visual synthesis, and multi-level quality curation. Ego2Robot supports both curated datasets and in-the-wild videos, producing 18,561 hours of robot training data spanning 15 robot morphologies, making it the largest ego-to-robot dataset to date. To evaluate generalization, we extend RoboTwin2.0 with disentangled perturbation axes covering visual appearance, scene layout, embodiment morphology, and task semantics. Experiments show that joint pretraining on Ego2Robot-synthesized and robot data consistently improves out-of-distribution generalization across multiple perturbation types, with benefits validated on real-robot deployment. Project page: https://www-ye.github.io/ego2robot_blog/
ELECTRIC: Evidential Learning-Enhanced CT Reconstruction via Iterative Correction
ELECTRIC：通过迭代校正增强证据学习的CT重建
Wang, Ge
Abstract
Here we introduce ELECTRIC (Evidential Learning-Enhanced CT Reconstruction via Iterative Correction), a physics-guided Bayesian formulation. An evidential neural network provides an image proposal and an error-predictive epistemic-uncertainty surrogate. The latter is converted into an adaptive precision field and inserted into a Poisson-weighted MAP update. The resulting image-evidence-precision-reconstruction loop treats prior confidence as a learned state variable of iterative reconstruction. In addition to the formulation and theoretical analysis, we report two simulation studies on image slices from the AAPM Mayo Clinic Low-Dose CT dataset: a mechanism-validation pilot using transparent surrogate estimators, and a feasibility study in which a trained Normal-Inverse-Gamma evidential network drives the full closed loop. On held-out patients, the learned prior mean reduces reconstruction error by roughly 70 percent relative to filtered back-projection, the learned epistemic uncertainty is error-predictive and supports selective trust, and the physics-guided update restores measurement consistency while the adaptive-precision reconstruction matches or exceeds a validation-tuned fixed prior and remains markedly more robust to prior-strength misspecification. Together these results demonstrate the complete ELECTRIC closed-loop pipeline, while identifying formal uncertainty calibration and joint training as the principal directions for future work.
Noise-Robust Conditional Flow Matching: Generating Clean Samples from Noisy Datasets
抗噪声条件流匹配：从噪声数据集中生成干净样本
Urbański, Adrian, della Maggiora, Gabriel, Yakimovich, Artur
Abstract
Generative models learn the statistical properties of their training data, so high-quality generation depends on clean and representative datasets. In scientific imaging, acquisition often yields noisy measurements, while collecting clean references can be costly, impractical or even unattainable. Training directly on these measurements results in a model that reproduces the corrupted data. This can be circumvented by learning the clean population distribution directly from the noisy data. Conditional flow matching (CFM) combines a simple regression objective with stable training, efficient sampling, and strong image-generation performance, making it a natural framework for this setting. We introduce Noise-Robust Conditional Flow Matching (NR-CFM), an unconditional generator that learns from one corrupted observation per image. NR-CFM provides a closed-form clean endpoint correction for additive white Gaussian noise and learns a data-driven correction for general Gaussian corruptions with more complex covariance structure. Across the evaluated corruption settings, NR-CFM outperforms NR-GAN in most cases and remains competitive with Ambient Diffusion in the high-noise regime. We further evaluate NR-CFM on scientific data at signal-to-noise ratios as low as $0.001$, where it generates plausible particle images from severely corrupted measurements.
Remote photoplethysmography (rPPG) enables non-contact heart-rate estimation from facial videos, but its weak physiological signal is easily corrupted by motion, illumination changes, occlusion, skin-appearance variation, and device noise. Existing rPPG methods typically rely on a single model to directly predict heart rate or recover pulse waveforms, while different strong estimators may produce conflicting yet individually plausible candidates for the same video. To resolve these conflicts, we propose PhysAgent, an inference-time multi-agent candidate-verification framework. Unlike direct prediction approaches, PhysAgent neither trains a new base rPPG model nor asks Multimodal Large Language Models (MLLMs) to output heart rate directly. In contrast, it treats outputs from multiple base estimators as physiological hypotheses to be verified and uses a lightweight 4B MLLM, Qwen3-VL-4B, to drive multi-agent reasoning over video conditions, signal reliability, and candidate disagreement. A deterministic physiological verifier checks the fusion proposal, and a reproducible numerical fusion process produces the final heart rate. Experimental results on multiple public rPPG benchmarks show that PhysAgent improves fusion stability and reliability across different datasets and source-domain settings, while avoiding the irreproducibility and physiological inconsistency of direct MLLM prediction or unconstrained ensemble fusion. The code will be released soon.
Construction-safety models must handle concrete deployment risks, such as a worker standing near a scaffold edge without guardrails, rather than only recognize common objects in curated images. Yet real inspection archives are redundant, long-tailed, and collected across changing sites and months. We introduce SafeBuild-Bench, a metadata-driven benchmark for evaluating multimodal large language models on construction safety under realistic temporal and site variation. It is mined from 100K+ industrial image-text records and contains 3,314 task instances from over 3,000 expert-verified images, covering multiple-choice hazard identification and free-form hazard description. To make expert verification scalable, we develop GEMS, a graph-enhanced multimodal selection pipeline that combines a proxy-model confusion signal with graph-based diversity to identify informative candidates from redundant streams. On public instruction-tuning data, GEMS-selected subsets preserve robustness-oriented performance under small data budgets. On SafeBuild-Bench, current MLLMs remain far from reliable construction-safety understanding, with the best overall score near 60. We release the benchmark, evaluation scripts, and GEMS codebase at https://github.com/safebuild/gems.
The rapid emergence of 3D CT foundation models has opened new avenues for predictive modeling from CT imaging, offering a compelling alternative to traditional radiomics which is known to suffer from reproducibility issues and sensitivity to acquisition protocol variations. Yet, as these models grow in availability, a critical need arises to evaluate how well their learned representations generalize across diverse clinical settings and whether adaptation to specific downstream tasks is necessary to unlock their full potential. To address these questions, we benchmarked several 3D CT foundation models for predicting recurrence-free survival in head and neck cancer across two public datasets totaling 3,644 patients, evaluating various adaptation strategies and modality fusion mechanisms. Our findings reveal persistent difficulty in identifying features that generalize consistently across different imaging distributions, as evidenced by significant performance drops on external validation cohorts. Ultimately, the integration of imaging features with clinical data remains the most accurate approach for prognostic prediction, though achieving universal generalization across varied clinical contexts continues to represent a substantial challenge for the current generation of models.
Chinese Translation
3D CT基础模型的快速出现为基于CT影像的预测建模开辟了新的途径，提供了一个引人注目的替代方案，克服了传统放射组学在可重复性问题和对采集协议变化的敏感性方面的不足。然而，随着这些模型的可用性不断提高，迫切需要评估它们所学习的表征在不同临床环境中的泛化能力，以及是否需要针对特定下游任务进行适应，以充分发挥其潜力。为了解决这些问题，我们对几种3D CT基础模型进行了基准测试，以预测头颈癌患者的无复发生存期，涵盖了两个公共数据集，总计3,644名患者，并评估了各种适应策略和模态融合机制。我们的研究结果显示，在不同影像分布中持续存在识别特征的一致性泛化困难，外部验证队列的显著性能下降证明了这一点。最终，将影像特征与临床数据结合仍然是进行预后预测的最准确方法，尽管在不同临床背景下实现普遍泛化仍然是当前模型面临的重大挑战。
Recent advances in quantum computing are opening new possibilities for Earth Observation (EO) data analysis. Quantum machine learning (QML) approaches offer novel ways to process information by exploiting quantum phenomena such as superposition and entanglement. These capabilities have motivated the exploration of whether quantum-enhanced models can address long-standing challenges in satellite remote sensing, where complex spectral and spatial signals often require sophisticated feature extraction. Among various fields of application, EO data allow the global monitoring of volcanic clouds and are crucial for aviation safety, hazard assessment, real-time eruption response, and evaluation of volcanic impacts on climate. Yet accurate detection of volcanic clouds remains difficult due to their similarity with meteorological clouds, the variability of eruption signatures, and the coarse spectral sampling of geostationary sensors. In this work, the potential of hybrid quantum convolutional neural networks (QCNNs) for the classification of satellite images containing volcanic clouds was investigated. These architectures integrate quantum computational layers into a classical convolutional framework. Two QCNN variants (with 2 and 4 qubits) have been considered to evaluate their ability to classify a dataset of SEVIRI images, including scenes with volcanic clouds (composed of ash, $SO_2$, or mixed components) as well as non-volcanic backgrounds. Finally, the performance of the hybrid QCNN models was compared with that of purely classical architectures.
Beyond Random Partitioning: Unsupervised Spatio-Temporal Stratification for Cohort Balancing in Longitudinal Medical Imaging
超越随机分割：用于纵向医学影像中队列平衡的无监督时空分层
Liu, Qinghui, Ottesen, Jon André, Bjørnerud, Atle, Emblem, Kyrre Eeg
Abstract
Rigorous dataset partitioning is a foundational, yet frequently overlooked, prerequisite for reliable deep learning in longitudinal medical imaging. Naively shuffling small clinical cohorts routinely introduces covariate shifts and temporal sampling imbalances across training, validation, and test subsets, exposing downstream models to out-of-distribution evaluation. We address this vulnerability with an auditable Tripartite Dataset Analytics Framework that systematically characterizes spatial grid integrity, multi-parametric intensity fingerprints, and longitudinal temporal trajectories, quantifying the heavy-tailed feature dispersion and irregular, episodic sampling intervals typical of real-world clinical cohorts. Building on this characterization, we formalize an unsupervised spatio-temporal cohort-balancing standard operating procedure (SOP) that combines elbow-optimized K-means clustering over a standardized, six-dimensional joint intensity-temporal feature space with intra-cluster proportionate stratified sampling. On a longitudinal, contrast-enhanced $T1$-weighted brain MRI cohort (N=149), the protocol reduces the maximum cross-subset intensity bias from 34.1% under conventional random shuffling to under 2.1%, while aligning longitudinal follow-up intervals closely around the population mean. Monte Carlo stress testing across ten random seeds and three split configurations confirms that this alignment remains tightly bounded, in clear contrast to the substantial variability of random partitioning. The resulting protocol offers a reproducible, generalizable procedure for cohort engineering in variable-length longitudinal clinical imaging workflows.
This paper presents a multimodal machine-learning framework for calibration monitoring, quality assessment, and adaptive acquisition support in archaeological digitisation workflows. The proposed approach operates across photogrammetric 3D reconstruction, hyperspectral imaging, X-ray fluorescence spectroscopy, and Raman spectroscopy through a unified pipeline combining deterministic quality indicators, statistical feature representations, machine-learning classification, anomaly detection, and explainable artificial intelligence (XAI). Rather than replacing instrument-level calibration, the framework introduces an additional algorithmic layer that evaluates whether acquisitions are statistically consistent, physically plausible, and suitable for downstream multimodal integration. For each sensing modality, acquisitions are represented through structured feature spaces encoding geometric, spectral, spatial, and statistical properties. These representations are used to identify degradation patterns such as reconstruction artefacts, illumination inconsistencies, spectral distortions, detector instability, baseline fluctuations, and low signal-to-noise conditions. Supervised and unsupervised learning methods are combined with XAI techniques to support both automatic discrimination between acceptable and problematic acquisitions and interpretation of the underlying causes of degradation. The framework additionally supports adaptive feedback and resource-aware acquisition strategies by linking feature-space deviations to acquisition-level corrective actions. Experimental results obtained on multimodal archaeological datasets demonstrate that the proposed methodology captures meaningful acquisition variability and enables robust quality assessment across heterogeneous sensing modalities.
Accelerated magnetic resonance imaging (MRI) aims to recover the k-space signal from acquired measurements, where accurate estimation of missing samples is essential for high-fidelity reconstruction. Existing k-space reconstruction methods estimate missing samples through interpolation operators or structure priors defined on discrete sampling grids. Although these formulations effectively exploit local interpolation relationships and global k-space redundancy, they reconstruct only discrete frequency coefficients and therefore do not explicitly model the underlying continuous signal. To overcome this limitation, we propose K-space Gaussian Representation (KGR), the first explicit continuous representation formulated directly in the native k-space domain. Rather than estimating unknown samples on discrete grids, KGR parameterizes the continuous signal using Gabor-Gaussian primitives with shared spatial geometry, yielding a compact representation that naturally preserves inter-coil correlations. Because unconstrained continuous fitting does not necessarily satisfy the intrinsic structural properties of multi-coil signal, the estimated representation is projected onto a low-rank manifold to enforce the algebraic constraints arising from smoothly varying phase and coil redundancy. A frequency-adaptive fitting strategy accommodates the heterogeneous characteristics of different k-space regions. Comprehensive validation across multiple datasets and sampling schemes shows consistent improvements over representative reconstruction baselines in both quantitative metrics and visual quality. These results suggest that explicit continuous parameterization of native k-space provides a principled framework for integrating continuous signal modeling with structured low-rank reconstruction.
Multimodal large language models (MLLMs) increasingly support high-stakes decision making by combining complementary information from images and text. While existing explainability methods identify influential image regions or text tokens, they cannot answer a fundamental question: which modality drives a prediction? Consequently, a model may produce the correct output while relying on the wrong source of evidence, masking shortcut learning and unsafe reasoning. We formulate modality attribution as a complementary explainability objective for multimodal foundation models and propose Counterfactual Modality Attribution (CMA), the first framework for quantifying modality-level contributions in MLLMs. CMA generates image-only, text-only, and joint multimodal counterfactuals using coupled diffusion priors and converts them into principled modality attribution scores through a cooperative game-theoretic formulation based on Shapley values. We evaluate CMA on controlled synthetic benchmarks with known ground-truth modality reliance and on a real-world multimodal clinical dataset. CMA correctly identifies the decision-driving modality in 98% of controlled cases and consistently outperforms baselines, revealing failures of cross-modal reasoning that remain invisible to predictive accuracy alone. Our results establish modality attribution as a complementary dimension of explainability beyond feature attribution, providing a principled framework for auditing multimodal foundation models in safety-critical applications.
Visual-token pruning is usually judged by answer quality at a fixed retention budget. For text-rich multimodal large language models (MLLMs), this protocol can miss a distinct failure: an answer remains correct even when no retained token is locally traceable to the small OCR region that supports it. We turn this blind spot into an evidence-risk audit that couples answer behavior with geometric token-origin provenance, interventions, and realized cost; transparent training-free selectors isolate controlled operating points. On locked image-disjoint confirmation, Qwen Target at 30% retention has observed accuracy 0.786 versus 0.783 for Full (paired image-cluster difference +0.003, 95% CI [-0.014, +0.020]), yet same-budget Target, Random, and Grid retain sharply different positive-support coverage: 0.620, 0.270, and 0.318. Across Qwen3-VL-8B, LLaVA-1.5-7B, and InternVL3.5-8B, matched controls, interventions, detector tests, and external methods reveal model-specific quality-risk-traceability frontiers that accuracy alone does not expose. Materialized prefixes yield up to 4.32x batch-prefill speedup and 76.4% lower incremental peak memory; full-validation TextVQA and DocVQA further show that favorable target-verification points do not imply task-general compression. Visual-token pruning should therefore report surviving spatial provenance and realized cost alongside quality and compression.
Din, Saif U, Hussain, Muhammad Ahsan, Timofte, Radu, Ignatov, Dmitry
Abstract
Deploying convolutional neural networks generated by large language models (LLMs) on real mobile hardware requires more than GPU validation accuracy: INT8 TensorFlow Lite export, delegate selection, and on-device latency jointly determine whether a model is usable. We present an automated mobile deployment pipeline that closes the loop from QLoRA fine-tuning of an architecture-generating LLM through GPU evaluation, INT8 export, and physical-device benchmarking to gated augmentation of the training corpus. The pipeline is fully scripted and runs cycle-by-cycle without manual intervention, with resume support after interruptions. We evaluate the same frozen protocol on two benchmarks, CIFAR-10 and CIFAR-100, on a Samsung SM-P613 tablet (seed 42, 20 models per cycle, cycles 0-6). On CIFAR-10, cycle 1 is gate-accepted and improves the mobile deployment score approximately 25.6x over the baseline with a mean quantized accuracy of 46.9%; later cycles raise GPU accuracy but fail the non-decreasing mobile gate. On CIFAR-100, the pre-QLoRA baseline retains the best mobile score; iterative rounds improve GPU accuracy (up to 26.2%) yet cannot surpass cycle 0 on-device, and the training pool stalls at 19 examples after the first accepted round. Together, the two studies show that closed-loop GPU fine-tuning does not guarantee monotonic mobile gains, especially on harder classification tasks, and that multi-dataset, on-device measurement is needed to stress-test deployment objectives. We release per-cycle metrics with 95% confidence intervals, all figures, and complete reproduction commands.
LeapTalk: Breaking the Latency-Quality Trade-off in Talking Head Generation
LeapTalk：打破谈话头生成中的延迟-质量权衡
Zhang, Rongxiang, Liu, Songhua
Abstract
Long-form and real-time talking-head generation remains challenging due to a latency-quality trade-off: inefficient multi-step diffusion prohibits streaming generation, whereas real-time autoregressive approaches suffer from error accumulation and identity drift. To address this drawback, we propose LeapTalk, a novel framework that achieves stable and real-time talking-head generation with a single forward step, scaling to arbitrarily long videos. At the heart of our approach lies a single-step bridge distillation scheme. On the one hand, departing from the conventional noise-to-data paradigm, we introduce a data-to-data transport formulation based on a Brownian bridge. Anchored by a persistent reference, this strategy effectively mitigates identity drift and enhances long-term temporal stability. On the other hand, to enable smooth knowledge transfer from a pre-trained diffusion teacher to the student bridge model, we explore a heterogeneous distillation framework with an SNR-aligned time transformation $\Phi(\tau)$, which bridges the functional discrepancy between the two models. Moreover, we propose an audio-driven classifier-free guidance mechanism to maintain fine-grained lip synchronization under extreme step reduction. Extensive experiments demonstrate that our method achieves high-fidelity and temporally consistent video generation with only 1 step at up to 200 FPS, significantly outperforming existing approaches in both efficiency and stability. Project Page: https://zhangrongxiang.github.io/leaptalk-page/
Beyond Edge Maps: Wavelet-Domain Conditioning for Multi-Adapter Map-to-Satellite Diffusion
超越边缘图：小波域条件下的多适配器地图到卫星扩散
Prasain, Arisha
Abstract
Commercial mapping partnerships are often unavailable in low-resource regions, leaving satellite basemaps stale and motivating synthesis of satellite imagery from independently maintained cartographic data. Existing ControlNet-based diffusion methods typically condition on structural signals like edges or segmentation extracted from the target image itself, assuming the imagery already exists and limiting their use exactly where synthesis matters most. Map-conditioned alternatives add cues like edge detection but omit frequency-domain structure. We propose a ControlNet-based diffusion framework conditioned only on cartographic sources obtainable independently of the target imagery: OpenStreetMap (OSM) raster maps and their stationary wavelet transform (SWT) subbands, a conditioning signal previously unexplored for map-to-satellite diffusion. Two ControlNet adapters, trained separately on the map and wavelet representations atop a frozen Stable Diffusion backbone, are fused via MultiControlNet, jointly drawing on spatial structure and frequency detail without retraining a multi-input model. We evaluate on a new paired map-satellite dataset curated for Nepal, a data-scarce, topographically diverse region, alongside the Pix2Pix maps-satellite benchmark. Combined conditioning wins six of eight metric-dataset comparisons -- SSIM and PSNR on both datasets, plus LPIPS (both Alex and VGG backbones) on ours and ties map-only on both Pix2Pix LPIPS backbones while still edging past wavelet-only there. Wavelet-only takes the lowest FID on both datasets, matching the tradeoff between per-image fidelity and distributional realism. We treat this gap cautiously given our modest test-set sizes and FID's known small-sample bias.
We present PixCell, a neurosymbolic system in which multimodal agents convert a visually presented photonic component into a parametric program over a small domain-specific language (DSL) of geometric primitives. A system enabling deterministic visual verification renders evaluation asymmetrically cheaper than the generation attempt. While models using multi-seed sampling and iterative revision reach a mean best-turn IoU of only 0.416, multimodal agents through PixCell's interface and verifier consistently exceed 0.9 mean IoU, with scores reaching 0.974 and 0.955 across eight component targets while also satisfying source contracts. These results demonstrate that frontier multimodal agents can reliably understand and render executable parametric representations from visual targets. Using these live parameters, cross-stack studies on an interferometer reconstruct primitive programs that satisfy an 8.0 nm free spectral range target and the original footprint constraint on modeled 220-nm SOI, 400-nm SiN, and 400-nm TFLN stacks. PixCell further carries a paper-derived splitter from visual reconstruction through SOI full-wave simulation, producing symmetric propagation and balanced outputs. Finally, the same executable verifier supplies a training reward and dataset used to train a Qwen3.6-35B-A3B model with LoRA and GRPO without supervised demonstrations. On eight training-excluded paper figures, its mean champion IoU rises from 0.422 after eight initial attempts to 0.491 after three verifier-guided revision rounds. These results therefore establish a controlled framework for measuring, retargeting, and improving visual-to-parametric photonic component design.
Brain tumor diagnosis is a time-sensitive process in which patients may wait weeks for a finalized pathology report. This problem motivates automated systems that classify tumor subtype from multimodal inputs. This paper details the DS@GT ARC team's work for ImageCLEFmed MEDIQA-CORE 2026 Task~1, Brain Tumor Subtype Classification. The task evaluates three glioma classification problems: Level-1 Molecular Type, LGG vs HGG, and WHO Grade. We combine pre-extracted MRI (NeuroVFM) and histopathology (Prov-GigaPath) embeddings with free-text radiology reports. Our team explored two trimodal fusion architectures, two report encoders (RadBERT and Llama-3.1-8B-Instruct), and a biologically motivated post-processing stage. We achieve a mean macro-F1 of 0.801 under the Fully Multimodal condition, exceeding the organizers' baseline of 0.796 and ranking second among the teams whose code passed verification. Additional evaluation across modality-dropping conditions shows that this advantage depends heavily on the availability of the histopathology modality, and that our system falls behind the baseline when modalities are missing. Our code is available on GitHub at https://github.com/dsgt-arc/imageclef-mediqacore-2026.
DODA: A Database of Datasets for Aesthetics Research
DODA：美学研究数据集数据库
Koßmann, Lisa, Bartho, Ralf, Redies, Christoph, Wagemans, Johan
Abstract
With rapid growth in the fields of empirical and computational aesthetics we have seen a vast increase in large image datasets annotated for aesthetics. As the image databases differ widely in many respects (e.g., different standards for annotation), it can be tedious to find the dataset that fits one's research needs best. The absence of a centralized open-science search system causes additional problems. Currently, researchers typically share dataset links in papers or on diverse platforms like OSF, GitHub or Dropbox. Manually searching for details like image quality and content often requires downloading all datasets. Therefore, we present the Database Of Datasets for Aesthetics (DODA), an intuitive Web application in which researchers can browse all important datasets for aesthetics research. DODA provides general information about these datasets (size, resolution, type of annotation, number of annotators, etc.) and for many of them also precomputed quantitative image properties. We discuss relevant criteria for selecting a suitable dataset with DODA and illustrate the benefits of reusing datasets. Our approach facilitates collaboration across the fields of empirical and computational aesthetics. Keywords: empirical aesthetics, computational aesthetics, machine learning, image annotation, quantitative image properties, Open Science
Chinese Translation
随着实证美学和计算美学领域的快速发展，我们见证了大量用于美学标注的大型图像数据集的迅速增加。由于这些图像数据库在许多方面存在显著差异（例如，标注标准不同），因此寻找最符合研究需求的数据集可能会变得繁琐。缺乏一个集中式的开放科学搜索系统进一步加剧了这一问题。目前，研究人员通常在论文中或在OSF、GitHub或Dropbox等不同平台上分享数据集链接。手动搜索图像质量和内容等细节通常需要下载所有数据集。因此，我们提出了美学数据集数据库（Database Of Datasets for Aesthetics，DODA），这是一个直观的Web应用程序，研究人员可以在其中浏览所有重要的美学研究数据集。DODA提供了关于这些数据集的一般信息（大小、分辨率、标注类型、标注者数量等），并且对于其中许多数据集，还提供了预计算的定量图像属性。我们讨论了使用DODA选择合适数据集的相关标准，并说明了重用数据集的好处。我们的方法促进了实证美学和计算美学领域之间的合作。关键词：实证美学，计算美学，机器学习，图像标注，定量图像属性，开放科学
Pretrained video diffusion models can act as renderers when the desired scene state is already specified by an animated mesh, a camera trajectory, and a reference image. This 4D generative rendering setting raises a representation question: what image-format condition lets a video backbone obey both camera motion and scene-internal animation? We propose DAR, a reference-guided renderer that extends Wan2.2 camera control from Pl\"ucker rays alone to a joint camera-plus-geometry interface. DAR projects a neural 4D G-buffer (tracking, world position, and normal) from the animated mesh and injects it through a widened control adapter while preserving the pretrained image-to-video prior. The central design choice is the pair of tracking and world position. Tracking identifies the persistent surface element that should carry appearance; world position gives its current scene-coordinate state; normal supplies local shape. Depth plus calibrated rays can recover 3D in principle, but depth is a camera-dependent chart in which camera and object motion are mixed. On the 68-case DAR-4D benchmark, LoRA DAR reaches PSNR 23.22, SSIM 0.895, and LPIPS 0.134, improving over off-the-shelf Wan2.2-Depth by 1.54 dB PSNR; a full fine-tune reaches PSNR 25.36 and SSIM 0.917. Matched ablations show that replacing world position by depth reduces PSNR by 1.26--1.55 dB at every checkpoint, supporting tracking+world-position correspondence as a practical 4D rendering condition.
Current deep learning-based character vision studies, e.g., text recognition, character image denoising, and historical text completion, are offering new solutions for learning, managing, and utilizing character resources. However, the performance of these studies peaks only with large and balanced datasets, which is a rarity with real-world character datasets, especially for logographic character languages, e.g., Chinese. The imbalance in data distribution of logographic characters is a common issue due to differences in character usage frequency and new characters being continuously created. In this paper, we propose a novel method for logographic character recognition, which introduces a multi-modal learning approach using visual semantics and contextual semantics of characters. A novel pre-training strategy is designed to enhance deep visual representations, especially for datasets suffering from issues of imbalanced and rare instances, by extracting the contextual semantics of each character from the corresponding language models. We conduct experiments across various datasets to evaluate our character recognition method and further validate the contrastive pre-training strategy by several downstream tasks. Experimental results demonstrate the superiority of our method compared to state-of-the-art methods.
Vision-language models (VLMs) are increasingly being evaluated for medical imaging, but many available benchmarks emphasize disease classification, report generation, or broad visual question answering rather than the spatial and anatomical reasoning required for radiology. We developed the Spatial Perception and Anatomical Reasoning in Clinical Radiology (SPARC-Rad) Benchmark, a manually curated multimodal benchmark dataset and evaluation pipeline for assessing these capabilities in radiology VLMs. SPARC-Rad includes 300 image-question pairs derived from healthy control imaging studies in The Cancer Imaging Archive (TCIA), spanning CT, MRI, and radiography across the abdomen, chest, breast, neuro, and musculoskeletal categories. Radiology trainees manually designed and annotated questions to evaluate anatomical identification, localization, laterality, regional recognition, device identification, and inter-structure spatial relationships. The evaluation pipeline supports standardized prompting, structured output collection, response normalization, LLM-as-judge grading, human quality review, binary correctness scoring, and subgroup analysis by modality, anatomy, and reasoning type. SPARC-Rad provides a reusable framework for evaluating whether VLMs can provide reasoning for radiologic anatomy as a spatial system, supporting future model development, failure-mode analysis, and pre-deployment assessment.
What Carries the Signal in Pathology Foundation-Model Atlases? A Patient-Level Controlled Benchmark in Breast Cancer
病理基础模型图谱中信号的载体是什么？乳腺癌患者级别的对照基准研究
Ndubuisi, Chimdi Walter
Abstract
Pathology foundation models are reported to encode molecular programmes in tissue morphology, but the evidence is usually a cohort-wide ranked gene list rather than a prediction for a held-out patient. We rebuild such an analysis with the patient as the unit of evidence and ask which pipeline component carries signal. Across 11 frozen backbones, four pre-specified gene programmes and 285 TCGA-BRCA patients with paired slides and RNA-seq (44 cells; GroupKFold by patient, all preprocessing fitted inside the fold), ridge regression on mean-pooled embeddings predicts held-out programme scores at Spearman rho = 0.25-0.56, UNI2 strongest on all four (immune 0.556). A matched permutation null gives raw p ~ 1e-4 at 10,000 permutations for every cell; Holm-adjusted p = 0.0044. The signal is real but not uniformly morphological. Against competing models on the same patients and folds, embeddings beat tissue composition for ER/luminal, proliferation and immune (+0.280, +0.284, +0.479; p <= 0.003) but not basal, where compartment fractions alone reach 0.469 against the embedding's 0.493 (p = 0.77). Fifty-four interpretable cell-count features come within 0.043-0.085 on every programme. The geometric machinery contributes nothing measurable, and we identify why: the geodesic graph selects neighbours by Euclidean nearest-neighbour search and only reweights edges already chosen, so the topology is Euclidean by construction (Riemannian minus Euclidean = +0.0010, 95% CI [-0.0007, +0.0029]). Applied consistently the geometry is worse (-0.0117). Ridge regression beats the graph-and-metric decoder by +0.097 (CI [+0.069, +0.127]). The driver-count metric common in this literature is near-uninformative here: 91.8% of random six-gene panels recover >=5/6 drivers.
3D scene understanding requires reasoning about entity existence, spatial layout, and object relations, yet RGB images alone often provide insufficient 3D cues. Existing 3D-VLMs commonly rely on depth or 3D-position-aware inputs at inference time, introducing additional acquisition, reconstruction, or annotation costs that limit RGB-only deployment. We therefore study how training-time 3D evidence can be converted into spatial reasoning capabilities retained under RGB-only inference. We propose a privileged-evidence distillation framework that constructs a distillable teacher through a unified evidence interface and controlled residual injection, and transfers its knowledge to a deployable student receiving only RGB images and questions through logit and structured representation distillation. To avoid imitating teacher signals unsupported by RGB, we further introduce evidence-sensitivity-guided distillation, which uses corrupted evidence to identify highly evidence-dependent targets and down-weight their supervision. We also define a recoverability decomposition based on the matched baseline, teacher, and student, separating privileged gains into RGB-recoverable improvements and residual teacher advantages. Across four benchmarks, the teacher achieves the best result on 7 of 11 reported metrics among the compared methods. The RGB-only student outperforms its matched baseline on all 11 metrics, including gains of 10.4 ScanQA CIDEr and 19.1 Scan2Cap [email protected], without additional inference-time inputs. These results validate the effectiveness of training-time privileged 3D evidence distillation for both teacher performance and deployable RGB-only spatial reasoning. Separately, our matched baseline-teacher-student analysis characterizes privileged-gain transfer across evidence types and spatial skills.
Counting the Cost of War Under Satellite Embargo: Zero-Shot Estimation of Impacted Infrastructure
在卫星禁运下的战争成本计算：受影响基础设施的零样本估计
Ahmed, Saleh Sakib, Rahman, M. Sohel
Abstract
Rapid estimation of impacted structures - critical for conflict-zone humanitarian response - is frequently hindered by post-strike satellite data embargoes and imagery blackouts. We bypass this operational bottleneck by reframing impacted building mapping as a zero-shot geometric projection task on archival, pre-strike maps. Using coordinate and incident text from LiveUAMap and ArcGIS, Large Language Models extract weapon payloads (W) to project kinetic blast perimeters via Hopkinson-Cranz scaling (R_base = Z * W^(1/3)). To count exposed structures within these zones without post-strike imagery, we introduce two technical innovations: Adaptive Field-of-View to eliminate resolution (zoom) bias in 2D segmentation (SAMGeo), and 2.5D pseudo-height depth maps combined with segmentation masks to help Large Vision-Language Models (LVLMs) resolve overlapping, dense rooftops. Evaluated on 2026 Middle East conflict data, depth-augmented LVLMs dramatically outperform traditional segmentation in congested urban centers. This establishes a powerful hybrid paradigm for zero-shot crisis mapping: ultra-fast 2D segmentation for sparse rural zones, and depth-augmented LVLMs for dense urban environments.
RadPRISM: Schema-stratified radiology-report supervision for concept-disentangled image representations and visual grounding
RadPRISM：基于模式分层的放射学报告监督，用于概念解耦的图像表示和视觉定位
Drexel, Fabian, Fritzsche, Marlene, Stambollxhiu, Era, Kumpf, Miriam, Schmitzer, Lena, Schumann, Lea, Kahmann, Jannik, Puttkammer, Friedrich, Moll, Johannes, Lübberstedt, Jannik, Chaaben, Zeineb Ben, Narayanan, Anirudh, Bercea, Cosmin I., Ziegelmayer, Sebastian, Makowski, Marcus R., Rueckert, Daniel, Adams, Lisa C., Bressem, Keno K.
Abstract
Vision-language pretraining learns rich medical image representations from radiology reports, but previous model variants commonly operate within a single shared embedding space, so concept-level structure and interpretability must be recovered post hoc, limiting model transparency and, hence, clinical utility. We introduce RadPRISM, which makes a clinician-defined radiology schema a designated stratification axis: an on-premise large language model extracts per-concept text spans from free-text reports, and each clinical concept is aligned in its own dedicated visual subspace, turning concept stratification into direct, top-level alignment supervision. Instantiated on chest radiographs with a 19-concept schema over $203{,}602$ examinations from an internal multi-year archive, RadPRISM improved internal dataset zero-shot classification from $0.717$ (95% CI, $0.710-0.723$) to $0.868$ (95% CI, $0.863-0.872$) macro AUROC over a matched global-alignment baseline, performed on par with the purpose-built CARZero reference in external zero-shot classification while substantially outperforming it (up to 4.3-fold) in pointing-game visual grounding. In addition, a radiologist reader study demonstrated concept-stratified retrieval ability ($0.78$ macro retrieval correctness rate within rank 3), surfacing disentangled descriptive findings that report-level retrieval and fixed-label vocabularies cannot express. RadPRISM yields discriminative, spatially faithful, natively concept-stratified representations shaped by and transparently inspectable by clinicians.
Correspondence-based statistical shape modeling (SSM) is vital for population-level morphometric analysis, but conventional pipelines assume clean, fully registered surfaces. Real-world clinical photogrammetry scans are often noisy, partial, and cluttered, hindering the adoption of radiation-free surface imaging as a safe alternative to computed tomography (CT) for infant craniosynostosis. We present SCALP (Semi-supervised Correspondence via lAndmark Localization and sPectral warping), a two-stage framework that constructs consistent shape models directly from raw, imperfect surface scans. First, a semi-supervised Point Transformer leverages a small expert-annotated dataset alongside a large unlabeled cohort to accurately localize craniofacial landmarks with minimal annotation overhead. Second, these landmarks anchor a Laplace--Beltrami spectral deformation of an anatomical template, generating dense correspondences while naturally isolating the cranium from peripheral scanning clutter without manual preprocessing. Experiments on infant photogrammetry scans demonstrate that SCALP consistently outperforms state-of-the-art unsupervised point-cloud approaches, offering a clinically practical pathway toward objective, radiation-free head shape analysis.
Manifold-GS: Certified Hybrid Assets via Varifold-Conservative Gaussian Splatting
流形-GS：通过变体场保守高斯点云实现认证混合资产
Li, Boyang
Abstract
3D Gaussian Splatting (3DGS) gives high-quality novel-view synthesis, but its adaptive radiance primitives are not directly usable as structured assets: opacity is not an additive area measure, refinement can change the induced geometry, and watertight mesh extraction can hallucinate collision surfaces in unobserved regions. We introduce Manifold-GS, a certified hybrid asset layer for Gaussian scenes. The method separates appearance opacity from geometric quadrature mass, represents surface-like Gaussians as a discrete unoriented varifold, and exports only confidence-certified open surface patches while retaining uncertified content as residual splats. It provides refinement-conservative mass transport, local realizability diagnostics, source-preserving patch bindings, and conservative collision candidates. On three DTU scenes, a frozen asset benchmark shows zero patch-defined edit leakage, texture round-trip PSNR of 30.1/35.3/33.7 dB, and lower collision floater area than official 2DGS meshes on all scenes, with large gaps on two scenes. The result is a precision-coverage tradeoff rather than a universal reconstruction claim. External-region annotations, phantom-collision probes, and 5k-face simplification further support the certified asset interpretation, while RGB-only experiments show that local realizability is not sufficient for sparse-view surface identifiability.
Learning How Much, Not Just What: Cross-Patient Burden Order for CT Vision-Language Pretraining
学习多少，而不仅仅是学习什么：针对CT视觉-语言预训练的跨患者负担排序
You, Guoliang, Gong, Haifan, Chu, Xiaomeng
Abstract
Volumetric CT vision-language pretraining learns 3D representations from scan-report pairs, but global and anatomy-aware objectives supervise only correspondence: they establish what is present and leave how much unconstrained. Nothing separates a mild from an extensive case of the same finding along a consistent direction, so the graded burden language in reports collapses into a present/absent signal. Longitudinal supervision would supply this order, but patient-matched CT pairs are scarce at scale; cross-sectional cohorts already encode weak burden cues across different patients. We introduce Spectrum, an anatomy-conditioned framework that represents each study at whole-study and organ scopes. For each organ-mapped pathology, a rule-based scorer mines confidence-filtered lower-to-higher pairs of different patients, and Burden-Direction Alignment (BDA) aligns the pathology-conditioned image delta with the report delta at each scope, separating that direction from its reverse. Because the endpoints are different people, a target-conditioned aligner first makes them comparable, so the delta reflects burden rather than between-patient variation. BDA further separates the selected direction from its reverse, anchors it to the observed higher-burden endpoint, and enforces consistency across ordered triplets. Since every pair is drawn within a single pathology, BDA is designed to constrain intra-class structure that image-report contrast alone never touches. Spectrum attains 85.6 zero-shot AUROC on CT-RATE and 72.7 on external RAD-ChestCT, with consistent gains in linear probing and retrieval. Weak cross-patient order is thus a scalable complement to anatomy-aware correspondence, yielding burden-aware CT representations without longitudinal data.
Real-Time Visual Obstruction Detection in Surgical Augmented Reality
手术增强现实中的实时视觉障碍检测
Yang, Shih-Chin, Xiu, Yanming, Ye, Hanting, Chen, Qi, Rotondo, Elias, Gorlatova, Maria
Abstract
Surgical augmented reality (AR) can provide contextual guidance by overlaying virtual annotations, tool cues, and procedural information onto the surgical workspace. However, the virtual content may obstruct task-relevant real-world information, such as surgical instruments, and interfere with users' perception during time-sensitive surgical tasks. In this paper, we investigate visual obstruction detection for surgical AR and present a latency-aware pipeline that combines vision-language model (VLM)-based surgical-object recognition with segmentation-based obstruction reasoning. To reduce inference overhead, the system adopts a cascaded small-to-large VLM architecture with segmentation-guided early exiting and attention-based visual token pruning. The small VLM handles easy frames when its key-object prediction is supported by segmentation consistency, while difficult frames are forwarded to a large VLM with pruned visual tokens. We construct a pseudo-AR surgical obstruction detection benchmark by overlaying virtual content onto surgical-tool images and labeling whether the virtual content obstructs task-relevant instruments. Evaluation results show that the proposed system achieves 87.43% obstruction detection accuracy with an average end-to-end latency of 479 ms, reducing latency by 62.90% compared with a cloud large-model baseline. These results demonstrate the feasibility of latency-aware obstruction detection for surgical AR and motivate future work on dynamic surgical videos, multi-object scenes, and clinically grounded AR guidance content.
Vision-language models (VLMs) have emerged as a powerful framework for multimodal video understanding. However, they remain limited in the sign language translation task, where we identify a key failure mode of existing VLMbased translators: poor spatial-temporal visual grounding. In particular, we find that standard next-token cross-entropy does not directly provide signal for where and when the model should attend, causing models to overlook sign-relevant regions and frames. To address this challenge, we propose AttnSign, a VLM-based spatial-temporal attention steering framework for sign language translation. AttnSign first introduces spatial attention supervision for sign-relevant regions, such as face and hands, in each frame; then develops an RL-based motion-cadence steering method that encourages the model to explore and focus on sign-level keyframes. Experimental results on How2Sign and OpenASL benchmarks show that our proposed AttnSign consistently outperforms existing methods.
Vision-language models (VLMs) have recently emerged as a promising paradigm for end-to-end autonomous driving, enabling agents to map multimodal inputs and high-level navigation instructions directly to executable trajectories. However, in practice, these models exhibit a persistent command-following gap: predicted trajectories often show weak sensitivity to navigation commands, resulting in incorrect behavior at critical decision points. We identify this issue as a form of conditional policy collapse, where regression-based training under multimodal trajectory distributions encourages the model to rely on dominant visual priors while marginalizing the language-conditioned signal. To address this issue, we introduce a principled formulation of classifier-free guidance (CFG) for regression-based vision-language driving. We show that CFG can be interpreted as isolating the instruction-induced residual in the action space by contrasting conditional and unconditional predictions, thereby explicitly amplifying the effect of the navigation command at inference time. However, a standard two-pass CFG introduces prohibitive latency for real-time control and produces noisy instance-level guidance directions. Building on a mean-shift interpretation of CFG, we propose Latent-Centroid Steering (LCS), a single-pass guidance mechanism that replaces instance-level residuals with class-level latent shifts. By projecting conditional representations toward precomputed command-specific centroids, LCS performs class-level latent steering based on cluster geometry that is both more stable and computationally efficient. We demonstrate that LCS reduces inference latency by approximately 50% while achieving stronger command adherence and improved driving performance on both closed-loop (Bench2Drive) and open-loop (nuScenes) benchmarks. Code will be released.
Semantically Calibrated Evidence Composition for CT Vision-Language Learning
用于CT视觉-语言学习的语义校准证据组合
You, Guoliang, Gong, Haifan, Chu, Xiaomeng
Abstract
Learning transferable representations from CT-report pairs requires combining whole-volume context with anatomy-specific evidence. Existing methods typically emphasize either global CT-report alignment or fine-grained anatomy-level correspondence. Global alignment preserves broad study context but leaves the contribution of localized evidence implicit, whereas anatomy-level alignment explicitly grounds local findings but does not specify how independently represented evidence should interact, acquire study-level meaning, and contribute to a global CT representation. To address this gap, we propose SCOPE (Semantic Calibration Of comPosed Evidence), a framework for semantically calibrated evidence composition in CT vision-language learning. Under organ-specific report supervision, mask-guided queries with fixed anatomical identities extract context-aware organ evidence from shared, uncropped volumetric features, while an unrestricted global query retains access to whole-volume context. The global query then drives Local-Global Coupling to compose the organ evidence into a unified evidence representation. The composed evidence is subsequently calibrated using the diagnostic summary, providing study-level semantic supervision beyond local organ descriptions, and is finally integrated as a controlled residual into a context-preserving whole-volume representation aligned with the complete report. This progressive pathway connects localized evidence with study-level semantics without reducing the CT representation to a predefined set of organs. On CT-RATE and RadChestCT, SCOPE achieves macro AUCs of 85.0 and 72.2, respectively, outperforming the previous SOTA by 7.2 and 4.2, while also yielding substantial gains in linear probing and cross-modal retrieval. These results demonstrate the effectiveness of semantically calibrated evidence composition.
MDWD: A Street-Level Dataset for Municipal Solid Waste Detection in Dense Urban Environments
MDWD：用于密集城市环境中市政固体废物检测的街道级数据集
Lucas, Andrea Filiberto, Bugeja, Mark, Debono, Carl James, Seychell, Dylan
Abstract
Automated visual monitoring of urban environments is a growing Computer Vision research area, but municipal solid waste detection remains under-represented in dedicated benchmark resources. Existing waste-related datasets predominantly address individual litter detection, aerial imagery, or image-level classification, and none simultaneously provide street-level imagery, instance-level localization, and categorization of domestic waste streams within a structured municipal collection context. This paper introduces the Maltese Domestic Waste Dataset (MDWD), a street-level benchmark comprising 3,697 high-resolution images and 11,461 manually annotated instances across five domestic waste categories representative of Malta's municipal collection system. The dataset captures substantial variation in location, illumination, object scale, occlusion, and urban context. To establish reproducible baselines, a cross-architecture benchmark is conducted across multiple generations of the YOLO family and a transformer-based detector. On the test set, RF-DETR-M achieves the strongest overall performance with an mAP50 of 94.49% and an F1-score of 93.56%, whilst smaller-capacity variants maintain competitive accuracy at substantially reduced parameter counts. These results indicate that MDWD supports effective training across both compact real-time detectors and transformer-based models, establishing a benchmark for future research in vision-based municipal waste monitoring.
Vision foundation models, such as DINOv2, learn highly expressive representations but rely on massive, opaque architectures that demand substantial computational power and memory. To provide an interpretable-guided and efficient solution to this issue, we first propose a spectral analysis and new visualization technique for individual attention heads based on the Laplacian eigenvectors of their attention maps. Building upon recent observations regarding the block structure of Vision Transformers, we perform semantic clustering of attention heads and identify functional redundancies. Leveraging these insights, we introduce SAPER (Soft Attention PrunER), an end-to-end differentiable pruning framework based on the LapSum Soft Top-K approach. Extensive experiments on ImageNet-1K demonstrate that SAPER achieves a highly favorable accuracy-efficiency trade-off, outperforming the competitive RAPTOR baseline in FLOPs reduction while preserving strong classification performance.
A 3D CT scan entering a vision-language model produces a long sequence of visual tokens, often thousands to tens of thousands per volume, and this sequence must be compressed before a language model can consume it. Token compression is well studied in general vision, but little of it targets 3D CT specifically. A common baseline is grid average, which pools regular grid cells and can blend distinct anatomy, lesion, and air into one token. We present \textbf{ORCA} (ORgan-Centroid Aggregation), a token compressor for 3D CT. It merges adjacent tokens with organ guidance and adds a sinusoidal encoding of each region's centroid to preserve spatial layout. This preserves the anatomical information a downstream model needs. ORCA is training-free and plug-and-play, producing an adjustable token set without any model change or text query. We evaluate it across two datasets (CT-RATE and Merlin) and five encoders. The evaluation spans two task types: attribute prediction over five families (size, density, location, texture, and disease) and text generation (visual question answering and report generation). At matched token budgets, ORCA improves consistently over existing compression methods. It shrinks the visual context $64\times$ and its KV-cache $50\times$, and is $31\times$ faster to process each volume. Code released at https://github.com/renjie-liang/ORCA-3DCT.
The First AI Children Challenge aims to advance real-world applications of computer vision and AI in child healthcare, child education, and pediatrics. The 2026 CV4CHL edition featured the first track in this domain: Children Gait Visual Analysis. The main goal of Children Gait Visual Analysis is the fine-grained analysis of children's gait behaviors from keypoint sequences. This is still a big challenge for human action recognition. Experienced medical doctors can distinguish these subtle nuances, but none of the people test AI models in this domain. To bridge this gap, we introduce thousands of 2D children keypoint sequences walking around videos across various age groups of children (3-16 years old). There is a significant opportunity for batch analysis of these videos to provide clinically relevant insights into medical diagnosis. The Challenge will be launched with two problem tracks: Edinburgh Visual Gait Score (EVGS) Scoring and Classification of Gait Patterns in Bilateral Spastic Cerebral Palsy. Each track is chosen in consultation with board-certified pediatricians based on the value of potential solutions. With the first available dataset for such tasks and ground truth for each track, the challenge enabled participants to evaluate their solutions. Final rankings will be revealed after the competition concludes, fostering reproducibility and mitigating overfitting.
Artificial Intelligence for the Characterization of Particles and Fibers by Optical Microscopy
通过光学显微镜表征颗粒和纤维的人工智能
Sun, Simiao, Ng, Kenneth, Lee, Lynn, Harth, Astrid, Odate, Asami, Katsaggelos, Aggelos, Matito, Manuel Ballester, Eastaugh, Nicholas, Walton, Marc
Abstract
Optical microscopy of particle and fiber dispersions involves interpreting subtle visual cues influenced by specimen morphology, chemical composition, magnification, and illumination conditions. We introduce an artificial intelligence (AI) distillation framework that extracts semantically rich image embeddings from microscopy images using semantic anchors. A multimodal teacher combines each image's visual embedding with three text embeddings representing illumination modality, magnification, and specimen identity and morphology. Generated by LongCLIP's extended-context text encoder, this yields a 2304-dimensional block-structured teacher vector whose component blocks remain physically interpretable throughout training and inference. A student vision transformer (ViT) with a multi-layer perceptron (MLP) decoder is trained to reconstruct this teacher vector from the image alone, minimizing a mean absolute error (L1) loss that enforces coordinate-level fidelity to the teacher's block structure. A cross-entropy term over pseudo-classes derived from HDBSCAN clustering of the teacher embedding space acts as a collapse-prevention regularizer, enforcing inter-cluster separation without requiring contrastive negative mining. At inference, the student operates on image input alone, producing compact embeddings that recover the full semantic content of the teacher vector. The framework achieves approximately 80% pseudo-class validation accuracy and 75% Recall@1 on fine-grained specimen description labels under leave-one-out nearest-neighbor retrieval. These results demonstrate that semantic anchoring enables a vision-only student to acquire richer and more interpretable representations than image-only training, with direct applicability to retrieval, classification, and exploratory analysis of heterogeneous particle and fiber dispersions.
We introduce a new problem domain for human action recognition: the fine-grained analysis of children's gait behaviors from standard RGB video. We specifically target the ambulatory patterns of children aged 3-17 years. Such behaviors arise naturally in the diagnosis and treatment of several critical developmental and neuromuscular disorders, such as cerebral palsy and hemiplegia. Despite their clinical value, current 3D sensor-based gait analysis systems are expensive, intrusive, and often impractical for young subjects. To address this, we introduce a new dataset comprising over 1,100 high-frame-rate (60 FPS) video sequences from 110 subjects, accompanied by synchronized, anonymized pose sequences. In each session, the child performs a 5-second "walk-around" task, capturing the gait cycle from multiple viewpoints. Crucially, we demonstrate that current state-of-the-art approaches, including gait foundation models and Multimodal Large Language Models (MLLMs), fail to effectively resolve these clinical nuances. We identify the key technical challenges in analyzing these erratic and subtle motor patterns and describe a unified end-to-end framework for decoding fundamental components of pediatric gait. Through comprehensive experimental results, we demonstrate the potential of this dataset to drive novel research questions and establish a rigorous baseline for automated child gait assessment.
Chinese Translation
我们提出了一个新的研究领域用于人类动作识别：从标准RGB视频中对儿童步态行为进行细粒度分析。我们特别关注3至17岁儿童的行走模式。这些行为在诊断和治疗多种关键的发育和神经肌肉疾病（如脑瘫和偏瘫）中自然产生。尽管这些行为具有临床价值，目前基于3D传感器的步态分析系统价格昂贵、侵入性强，并且对于年轻受试者往往不切实际。为了解决这一问题，我们引入了一个新的数据集，包含来自110名受试者的1100多个高帧率（60 FPS）视频序列，并附有同步的匿名姿态序列。在每个会话中，儿童执行一个5秒钟的“走动”任务，从多个视角捕捉步态周期。至关重要的是，我们证明了当前最先进的方法，包括步态基础模型和多模态大型语言模型（Multimodal Large Language Models, MLLMs），未能有效解决这些临床细微差别。我们识别出分析这些不规则和微妙运动模式的关键技术挑战，并描述了一个统一的端到端框架，用于解码儿童步态的基本组成部分。通过全面的实验结果，我们展示了该数据集在推动新研究问题和建立自动化儿童步态评估的严格基准方面的潜力。
Depth estimation is a significant task for 3D perception in endoscopic surgeries. However, illumination interference and feature diversity in various endoscopic scenes are still challenges for generalizable depth estimation and ego-motion estimation. Based on this, a novel self-supervised framework, EndoMINI, is proposed for depth estimation in endoscopic scenes. Specifically, mixture of low-rank experts (MiLoRE) is proposed to perform parameter-efficient fine-tuning, which can also boost the model adaptation to scenes with different characteristics. Meanwhile, an intrinsic image alignment (IIA) is introduced into the training loss to alleviate the influence of light reflectance in endoscopy with a novel intrinsic image decomposition network. The proposed method is evaluated on SCARED datasets for supervised depth estimation, and two endoscopic datasets, Hamlyn and SERV-CT, for zero-shot depth estimation, compared with state-of-the-art works as well. The experimental results demonstrate outstanding performance of the proposed model and the effects of the main contributions.
Poplar: A Scalable Pipeline for Human-Centric Image Dataset Synthesis
Poplar：一个可扩展的人本图像数据集合成管道
Zou, Zhishan
Abstract
Recent image generators can synthesize convincing human-centric images, yet producing a useful collection remains different from producing a single successful image. A human-centric dataset must cover varied people and contexts, avoid implausible attribute combinations, preserve an everyday photographic character, and expose quality-control decisions at scale. We present Poplar, a reproducible Specify--Render--Inspect pipeline for human-centric image dataset synthesis. Specify samples structured attributes under commonsense constraints and verbalizes them as photography-oriented prompts. Render uses a realism-adapted image generator across composition-aware aspect ratios and retries obvious technical failures. Inspect applies a single structured vision--language review to each candidate, preserving the original prompt while rejecting intrinsic image defects or material prompt mismatches. Using Poplar, we construct Poplar-9K: 9,401 curated human-centric image--text pairs retained from 11,765 reviewed candidates (79.9\% acceptance). We release the dataset together with the pipeline, configurations, immutable generation prompts, and auditable inspection records as a compact resource for building customizable human-centric collections.
Beyond Static Anchors: Bounded Prototype Conditioning for Language-Free Medical Anomaly Detection
超越静态锚点：无语言医学异常检测的有界原型条件化
Wan, Yibo, Cai, Jinyu, Seekiong-Ng
Abstract
Medical anomaly detection identifies abnormal images and localizes lesions under scarce supervision while generalizing across organs and modalities. Existing CLIP-based methods reduce annotation requirements through vision--language alignment, but their normal and abnormal references, whether text prompts or learned visual tokens, remain fixed across test images. Such static references may not transfer reliably to unseen targets in a cross-domain medical imaging scenario. To address this, we propose ReCAP, a language-free framework that replaces static anchors with input-conditioned visual prototypes. ReCAP re-centers separated normal and abnormal prototypes for each image through a bounded gated modulation, enabling query-adaptive anomaly scoring while constraining context-induced prototype drift. For the few-shot setting, we introduce a non-parametric normal-reference memory to preserve instance-level target-domain variation and complement the conditional prototype branch. Across six medical benchmarks, ReCAP achieves the best image-level AUROC on all zero-shot and 23 of 24 few-shot settings, and the best zero-shot pixel-level AUROC on all three segmentation datasets. Particularly, it reduces inference latency by over 70% compared to the fastest baseline, without text prompts or test-time gradient updates.
Unsupervised medical anomaly detection learns normal anatomical patterns from healthy training images and identifies deviations at test time. Reconstruction-based and diffusion-based methods commonly use the difference between an input image and its reconstruction as anomaly evidence. However, this residual can be ambiguous. Expressive models may preserve pathological structures, while benign anatomical variation, imaging noise, and acquisition differences may also produce large reconstruction errors. We propose discriminative mask-guided diffusion (DMD), a medical anomaly detection framework that complements residual-based localization with reconstruction-shift discrimination. DMD first learns a compact quantized latent representation of normal images. Localized masks then perturb selected latent regions, and a latent diffusion model reconstructs the perturbed representations. The resulting reconstructions are paired with their original normal images to define a self-supervised classification task. At inference, the classifier provides a learned image-level anomaly score, while the residual between the input and its diffusion-based reconstruction yields a pixel-level anomaly map. Experiments on five datasets spanning brain MRI, breast ultrasound, and chest radiography show that DMD achieves the best overall performance among the state-of-the-art baseline methods.
Structured Proxy Features for Multimodal NSCLC Survival Prediction from Pretreatment CT
用于多模态非小细胞肺癌生存预测的结构化代理特征：基于治疗前CT的研究
Nguyen, Huu Phong, Hossain, Delower, Saghapour, Ehsan, Sembay, Zhandos, Chen, Jake Y.
Abstract
Lung cancer results in roughly 1.8 million fatalities annually worldwide, with non-small cell lung cancer (NSCLC) comprising the majority of cases. Despite advancements in treatment, survival stratification remains challenging due to intratumoral heterogeneity inadequately captured by conventional descriptors. Standard radiomic and deep learning techniques regard imaging features as independent quantities, overlooking structured interactions between tumor characteristics. We evaluate whether structured proxy features can enhance multimodal NSCLC survival prediction by augmenting pretreatment computed tomography (CT) representations, radiomics, and clinical variables with six simulation-derived features designed to capture interactions between heterogeneity and morphology. A radiomic-parameterized cellular automaton generates growth-rate and necrosis-ratio proxy features from baseline CT by using entropy and sphericity to compute low-dimensional proxy parameters. The imaging backbone is a Transformer-based Masked Autoencoder (TMAE), which was chosen after a systematic evaluation with alternative encoders within the same pipeline and provides attention-based visualizations that highlight tumor regions receiving higher model attention. On the public Lung1 cohort (n = 390), the primary four-modality fusion attained a C-index of 0.641 (iAUC 0.731, log-rank p < 0.001). The primary result compares favorably with prior multimodal results on Lung1 (C-index 0.631; iAUC 0.592 [15]) under a comparable evaluation protocol, while a separate exploratory coefficient-optimization analysis achieved a best observed C-index of 0.662 (iAUC 0.748). These results indicate that, in addition to conventional radiomic, deep, and clinical representations within the Lung1 benchmark, simulation-derived proxy features may provide complementary predictive information within this fixed Lung1 benchmark.
Chinese Translation
肺癌每年在全球造成约180万例死亡，其中非小细胞肺癌（NSCLC）占大多数。尽管治疗有所进展，但由于肿瘤内部异质性未能通过传统描述符充分捕捉，生存分层仍然具有挑战性。标准的放射组学和深度学习技术将影像特征视为独立的量，忽视了肿瘤特征之间的结构化相互作用。我们评估结构化代理特征是否能够通过增强治疗前计算机断层扫描（CT）表征、放射组学和临床变量，来提升多模态NSCLC生存预测。我们设计了六个模拟衍生特征，以捕捉异质性与形态之间的相互作用。一个经过放射组学参数化的细胞自动机从基线CT生成生长率和坏死比率的代理特征，通过使用熵和球形度计算低维代理参数。影像基础架构是基于Transformer的掩码自编码器（TMAE），在对同一管道内的替代编码器进行系统评估后选择，并提供基于注意力的可视化，突出显示接受更高模型关注的肿瘤区域。在公开的Lung1队列（n = 390）上，主要的四模态融合达到了0.641的C指数（iAUC 0.731，log-rank p < 0.001）。主要结果与Lung1上先前的多模态结果（C指数0.631；iAUC 0.592 [15]）在可比评估协议下相比表现良好，而单独的探索性系数优化分析达到了观察到的最佳C指数0.662（iAUC 0.748）。这些结果表明，除了Lung1基准中的传统放射组学、深度和临床表征外，模拟衍生的代理特征可能在这一固定的Lung1基准中提供补充的预测信息。
3D Gaussian Splatting (3DGS) turns captured or generated imagery into photorealistic 3D world simulations that users can freely explore, yet these worlds remain silent. Because existing audio generation methods condition on a single image or viewpoint, their sound is tied to that observation and cannot stay consistent while a listener moves. We introduce the task of generating a spatially consistent soundscape for a given 3DGS world through auditory grounding, identifying which objects in the world should emit sound and anchoring each to a persistent 3D position, and present Scene2Sound, a training-free framework built on this grounding. From the input world alone, our pipeline selects viewpoints that jointly cover the scene, identifies sound-emitting objects with a vision-language model, and associates the multi-view detections into 3D instances through Gaussian set matching, which measures the overlap between the Gaussian sets that render each detection. Each source then receives generated audio that a standard object-based audio engine spatializes in real time at arbitrary listener poses. We further propose two spatial-consistency metrics, one testing whether rendered audio responds consistently to listener motion and one testing whether the claimed sources are supported by views held out from their placement. On a curated set of generated 3DGS worlds and on 3DGS scenes generated from real-world 360-degree captures, Scene2Sound preserves the audio quality of strong per-viewpoint baselines while remaining spatially consistent where per-viewpoint and single-panorama pipelines do not, and a user study confirms the perceptual benefit. Project page: https://masaki-lmd.github.io/scene2sound/.
Architectural drawings violate the usual assumption behind multi-view reasoning: plans and sections are cuts, while elevations are facade projections, so corresponding components change appearance in ways camera motion cannot explain. We introduce CrossProjection, an anchor-grounded diagnostic of whether vision-language models preserve component identity and externalize geometry across heterogeneous architectural views. It evaluates Matching, Registration, and Geometric Grounding through categorical judgments, candidate selection, and free point, line, and region localization. Across 23 real drawing sets and 1,954 categorical conditions per model, GPT-5.5 scores 82.4%, Qwen3-VL-32B-Instruct 62.2%, and GLM-4.5V 57.2%. A matched 200-target study crosses natural and vector-text-suppressed drawings with closed-candidate and free-geometry outputs. Candidate-supported performance is often higher, but free localization remains fragile: on natural drawings, point/region [email protected] is 54-76% for GPT, 8-10% for Qwen, and 14-36% for GLM; line endpoint [email protected] is 22%, 4%, and 0%. A coordinate grid recovers some GPT point/region precision but not lines. Three architecture-trained participants reach 87.3-93.3% categorical accuracy and 76-92% GT-region hit, supporting task feasibility rather than a population-level human ceiling. Because the categorical families do not form a same-item Matching-Registration contrast and interface controls alter multiple burdens, we avoid mechanistic claims. The supported conclusion is narrower: closed-choice or marked-element success does not entail reliable explicit geometric grounding. For drawing-guided CAD/BIM systems, categorical correctness should not be treated as evidence of candidate-free spatial reliability. Reusable on-sheet anchors, fixed-denominator scoring, and hash-locked artifacts establish an audit trail for this gap.
Accurate prediction of object trajectories during manipulation is essential for closing the perception-action loop. Progress is limited on two fronts: available datasets lack fine-grained language-to-motion annotations, and existing predictors either rely on privileged inputs such as video, depth, or CAD models, or recover motion from fully generated videos through costly, error-prone perception pipelines. We close the supervision gap with the MOVE dataset, 5,038 object-centric egocentric trajectories, each paired with a fine-grained natural-language instruction rather than a coarse verb-noun label. We further propose DreamTraj, which predicts a 6-DoF object trajectory from a single RGB image and a task instruction, requiring no video, depth, or CAD model at inference: rather than generating a video, it reads motion from the internal representations of a frozen image-to-video diffusion model at an early denoising step. A lightweight flow-matching Reader decodes query-key attention tracks and pooled hidden states into relative 6-DoF poses. To our knowledge, this is the first approach to directly decode object 6-DoF trajectories from intermediate video diffusion representations rather than generated pixels. DreamTraj sets a new state of the art on both translation and rotation against forecasters that consume multi-frame or privileged inputs, and runs 4.6x faster than generate-then-extract pipelines.
Single-photon avalanche diode (SPAD) cameras are promising for low-light and high-dynamic-range intensity imaging, but their practical use is limited by complex sensor-specific noise. Unlike time-correlated single-photon counting (TCSPC) systems, SPAD cameras record whether at least one detection occurred in each gate without photon timestamps in intensity imaging mode, making explicit noise decomposition difficult. We present a practical noise modeling and calibration framework for SPAD intensity denoising. Our forward model describes binary-frame accumulation with a Binomial observation process, models signal-independent dark noise as an exposure-dependent pure dark count term plus an exposure-independent dark-frame bias term, and incorporates pixel-wise response non-uniformity. We design a dedicated calibration procedure for the proposed model and use it to build a count-domain noise-synthesis pipeline for network training. For denoising, we further design a SPAD-specific dark-shading correction (SPAD-DSC) to remove most systematic noise before network training. We construct a real-world SPAD intensity dataset for testing. Experimental results demonstrate the superiority of the proposed noise model.
Can image-editing models discover visual rules in image space and complete problem-solving end-to-end? We tackle this question in the spirit of a human worksheet test (e.g., an IQ test), using problems that require models to read image-based instructions, recognize the problem, infer the answer, bind it to the correct destination, control output count, suppress unnecessary edits, and preserve the input and format. We introduce WISRD, a Worksheet Image-Space Rule Discovery benchmark with 11 core tasks under eight information conditions, spanning localized marking, filling, copying, counting, and no-edit suppression, together with four supplementary reasoning-stress probes for multi-step spatial manipulation, abstract pattern reasoning, logical inference, and constraint-based problem solving. We identify three key findings as follows. (i) Among the frontier image-editing models evaluated, Nano Banana Pro achieves the highest score. On the shared V0--V3 no-reference subset, the Auto-Strict proxy pass rates are 48.7% for Nano Banana Pro, 13.4\% for Qwen-Image-Edit, 11.5% for FLUX.2 Klein 4B API, 11.3% for FLUX.2 Klein 4B open-weight, and 0.0% for InstructPix2Pix. (ii) Analysis reveals that current image-editing models can partially rely on rendered in-image instructions even when the external prompt is absent or merely generic. (iii) In small supplementary diagnostics, Nano Banana Pro achieves 70.0% on 4-by-4 Sudoku and 22.9% on public RAVEN pattern-discovery items in image space.
Chinese Translation
图像编辑模型能否在图像空间中发现视觉规则并完成端到端的问题解决？我们以人类工作表测试（例如，智商测试）的精神来探讨这个问题，使用需要模型读取基于图像的指令、识别问题、推断答案、将其绑定到正确目标、控制输出数量、抑制不必要编辑并保持输入和格式的问题。我们引入了 WISRD，一个包含 11 个核心任务和八种信息条件的工作表图像空间规则发现基准，涵盖局部标记、填充、复制、计数和无编辑抑制，以及四个补充推理压力探针，用于多步骤空间操作、抽象模式推理、逻辑推理和基于约束的问题解决。我们识别出三个关键发现如下：(i) 在评估的前沿图像编辑模型中，Nano Banana Pro 达到了最高分。在共享的 V0--V3 无参考子集上，Nano Banana Pro 的 Auto-Strict 代理通过率为 48.7%，Qwen-Image-Edit 为 13.4%，FLUX.2 Klein 4B API 为 11.5%，FLUX.2 Klein 4B 开放权重为 11.3%，而 InstructPix2Pix 为 0.0%。(ii) 分析表明，当前的图像编辑模型即使在外部提示缺失或仅为通用时，也可以部分依赖渲染的图像内指令。(iii) 在小规模补充诊断中，Nano Banana Pro 在 4x4 数独上达到了 70.0%，在公共 RAVEN 图案发现项目的图像空间上达到了 22.9%。
Optical flow remains challenging in high-speed and low-light scenes, where the limited frame rate and sensitivity of conventional cameras lead to motion blur and underexposure. Single-photon avalanche diode (SPAD) cameras offer single-photon sensitivity and extremely fine temporal sampling. However, individual slices in these high FPS binary photon streams are too sparse for dense correspondence. Temporal aggregation can provide the spatial cues required by optical flow, but accumulating photons at fixed coordinates blurs moving structures. Motion-aware aggregation can reduce this blur, yet it depends on the flow being estimated. To address this dependency, we propose QuantaFlow, the first method for dense optical flow directly from SPAD streams. Instead of constructing a fixed input representation, QuantaFlow embeds SPAD representation construction into iterative flow refinement. At each iteration, the current flow coarsely aligns the slices within the source and target sub-streams. A photon-flux transformation then constructs multi-scale representations containing intensity and structural cues, while adaptive multi-scale fusion balances photon noise and residual motion blur at each pixel. The fused representations drive a feature-warping flow update, and the refined flow guides representation construction in the next iteration. We further construct a synthetic dataset for SPAD optical-flow training and evaluation. Experiments on the synthetic dataset and real-world SPAD data demonstrate the effectiveness and generalization of QuantaFlow.
Affordance grounding aims to localize the functional region for interaction, such as the handle to grasp or the button to press, rather than the whole object. This makes it more challenging than generic visual grounding because the target region is smaller, more ambiguous, and more dependent on task context, especially for compact vision-language models (VLMs) used in embodied settings. Recent sequence-level supervision and reinforcement learning improve coordinate prediction quality, yet compact autoregressive VLMs still lack reliable affordance-aware visual focus before coordinate generation: the model can produce better coordinate tokens while its cross-modal attention remains diffuse and weakly anchored to the true affordance evidence. To address it, we propose SpatialAfford, a two-stage framework that first aligns attention to the ground-truth affordance region through Spatial Attention Alignment (SAA), then refines coordinate prediction with Spatial-Aware GRPO. By explicitly teaching the model where to look before optimizing where to ground, SpatialAfford turns affordance grounding from a purely output-constrained objective into attention-grounded spatial reasoning. Across ShareRobot-Bench, ReasonAff, and PartAfford, SpatialAfford consistently improves affordance grounding, with a compact 4B model outperforming stronger 7B+ baselines.
RadYOLO: Computationally Efficient 3D Object Detection and Segmentation in CT and MRI
RadYOLO：在CT和MRI中计算高效的3D物体检测与分割
Geissler, Kai, Müller-Groh, Laurens, Meine, Hans
Abstract
Object detection and segmentation in three-dimensional medical images is a very active area of research. However, most proposed deep learning models carry a high computational cost, and only few aim to be broadly applicable, achieve high detection performance, and remain fast to execute on resource-constrained hardware. To address this gap, we present RadYOLO, a 3D extension of YOLO11 tailored to medical images. We compare it with nnU-Net and nnDetection on five datasets comprising CT and MRI data with varying object sizes and prevalence. RadYOLO's detection performance surpasses that of nnDetection on four of five datasets and is comparable on one. Compared to nnU-Net, RadYOLO performs better on lesion detection tasks, while nnU-Net excels at detecting large organs when precise localization is required. When rough object localization is sufficient, RadYOLO matches or outperforms nnU-Net on all five datasets. Regarding inference time, RadYOLO is 8-46x faster than nnU-Net on a GPU. Compared to nnDetection the speedup is even higher. When executed on a CPU, RadYOLO's inference runs within seconds (still faster than nnU-Net on a GPU) offering a significant advantage for clinical and edge-device deployment. RadYOLO repository: https://github.com/FraunhoferMEVIS/RadYOLO
Test-time Adaptation of Pelvic Bone Segmentation Models via Dynamic Reliability-Guided
基于动态可靠性指导的骨盆骨分割模型的测试时适应
Ren, Ling, Deng, Chao, Wang, Ziming, Xu, Yuecong, Zheng, Kai
Abstract
Reliable pelvic bone segmentation (PBS) from CT is essential for robot-assisted pelvic trauma surgery, yet deploying a source-trained model to a new hospital suffers from severe performance degradation due to cross-center domain shifts. While test-time adaptation (TTA) enables online model adaptation without accessing source data, existing methods show limited effectiveness for PBS, facing challenges including boundary degradation, anatomical inconsistency under domain shifts, and voxel-level class imbalance. To address these challenges, we propose a novel closed-loop dynamic Reliability-Guided TTA framework (ReGA) for PBS. Specifically, we introduce a pseudo-label reliability criterion termed Segmentation Inference Consistency Evaluation (SICE), which jointly measures region overlap and boundary deviation via dropout-based ensemble predictions. Based on SICE, a trust-weighted refinement module adaptively updates features to mitigate boundary errors in pseudo-labels. Furthermore, a confidence-weighted region-level contrastive learning strategy is proposed to enforce anatomical consistency. Finally, ReGA follows the teacher-student (TS) scheme to alleviate voxel-level class imbalance. Experiments on three heterogeneous 3D pelvic CT datasets demonstrate that ReGA consistently outperforms state-of-the-art TTA methods, enabling effective adaptation of the source-trained PBS model to unseen clinical domains. The code is available at https://github.com/Ren-ling/ReGA.
GuideGround: VLM-guided Semantic Understanding and Viewpoint-aware Reasoning for 3D Visual Grounding
GuideGround：基于视觉语言模型的语义理解与视角感知推理用于3D视觉定位
Wang, Yiwen, Deng, Yuyang, Long, Yihao, Zhao, Xi
Abstract
3D visual grounding aims to localize the target object in a 3D scene from a natural language query, requiring both fine-grained semantic understanding and viewpoint-dependent spatial reasoning. Existing methods typically formulate semantic understanding as an auxiliary closed-set object classification task and rely on multi-view feature aggregation for viewpoint reasoning, limiting semantic generalization and weakening viewpoint-specific evidence. We observe that vision-language models naturally provide complementary capabilities through open-vocabulary semantic understanding and global scene perception. Based on this insight, we propose GuideGround, a VLM-guided framework that complements rather than replaces task-specific grounding models by leveraging VLMs for semantic enhancement and viewpoint-specific hypothesis verification. Specifically, we replace auxiliary closed-set object classification with VLM-generated object semantic descriptions to enhance semantic understanding. Meanwhile, instead of directly aggregating multi-view representations, we preserve viewpoint-specific grounding hypotheses through per-view grounding and explicitly verify them using VLMs across candidate viewpoints. Extensive experiments on the ReferIt3D benchmark demonstrate that GuideGround consistently outperforms previous state-of-the-art methods. Comprehensive ablation studies further confirm the effectiveness of both the proposed semantic understanding and viewpoint reasoning strategies.
Infrared-visible image fusion under realistic degradation scenarios is a challenging task, as degradations not only cause a loss of reliable modality-specific information in observed images but also hinder the fusion process. Recent studies indicate that text can provide prior information about degradation characteristics, complementing the limited evidence available from corrupted input images and facilitating fusion. However, existing methods typically inject fixed global text representations into visual features, making it difficult for textual guidance to adapt to spatially varying degradations, local structures, and thermal saliency. To this end, we propose TGFusion, a text-guided latent-space flow matching framework that unifies degradation suppression and cross-modal fusion. TGFusion encodes task, degradation, and generation cues into structured prompts. To fully exploit these priors, we design a Prompt-conditioned Multi-stream Joint Flow Transformer that represents text as an independent semantic stream alongside fusion, visible, and infrared streams. Joint attention enables token-level bidirectional interaction and layer-wise updating among semantic and visual representations, allowing degradation semantics to dynamically guide reliable information selection and fusion latent generation. Extensive experiments on public benchmarks and complex degradation scenarios demonstrate that TGFusion achieves superior or competitive performance in perceptual quality, image naturalness, structural-detail preservation, and infrared-saliency retention, while remaining robust across diverse single and compound degradations.
Reinforcement learning (RL) for document parsing often relies on reference-based rewards rooted in edit distance (e.g., tree edit distance), yet it remains hard to optimize in the high-accuracy regime because such rewards become weakly discriminative: near-correct outputs receive very similar scores, providing limited learning signal for hard cases. We propose Step-Aware Annealing (SAA), a plug-and-play reward sharpening mechanism that progressively increases reward curvature during training, amplifying subtle quality differences among high-scoring samples while preserving stability in early learning. Built on SAA, we introduce DocPO, a document policy optimization framework with element-specific, reference-based rewards anchored by edit-distance signals: normalized string edit distance (NED) for text, tree edit distance similarity (TEDS) for tables, and a hybrid Rubric+edit reward for formulas. Experiments on OmniDocBench and DocElemHard show that SAA consistently improves GRPO-style RL across document elements over non-annealed rewards, without requiring additional human supervision for reward construction.
Hybrid-Domain Posterior Sampling for Inverse Problems via Latent Flow Matching
通过潜在流匹配的混合域后验采样用于逆问题
Wu, Hongjie, Xie, Yiping, Lv, Jiancheng
Abstract
Latent Flow Models have revolutionized compressed-space image synthesis, yet their application to high-fidelity inverse problems remains bottlenecked. In this paper, we trace this dilemma to a fundamental geometric limitation of pre-trained autoencoders, which we term \emph{First-Order Manifold Blindness}. Severe decoder compression (e.g., retaining only $\sim\!2\%$ of the original degrees of freedom) produces a rank-deficient Jacobian, rendering high-frequency measurement residuals in its orthogonal complement invisible to latent gradients even when the decoder can represent the target image. To overcome this bottleneck, we propose Hybrid-Domain Posterior Sampling (HDPS), a decoupled inference framework that disentangles physical measurement consistency from semantic prior modeling. HDPS diverges into the pixel space, leveraging Langevin dynamics to absorb precise orthogonal measurement gradients, and subsequently projects these structural corrections back onto the generative manifold. An optimization-based latent alignment is introduced to filter pixel-space artifacts while avoiding the semantic drift of direct encoding. Extensive experiments on diverse inverse problems demonstrate that HDPS establishes a new state-of-the-art, successfully recovering the high-frequency structural precision that latent-only solvers inherently discard. The code is available at \href{https://github.com/74587887/HDPS}{https://github.com/74587887/HDPS}.
Tool-integrated vision-language agents have made remarkable progress on compositional and multi-step visual reasoning. Yet their outputs frequently exhibit unfaithfulness: the stated reasoning path diverges from the computation that actually produced the answer, undermining reliability in safety-critical applications. We present DiffuseAgent-MI, a self-evolving agent whose perceptual grounding is governed by a KL-minimal energy model over feature units, providing a distributional view of visual mechanistic interpretability. The agent learns an energy landscape that softly constrains generated samples to lie near the native prior conditioned on the chosen interpretable unit, closing the gap between the explanation and the internal representation. A verifier then supplies trajectory-level faithfulness rewards, and a repair branch re-conditions the energy when the verifier flags an unfaithful step. On GeoQA, SciVis, VQA-v2 and an in-house multimodal reasoning set, DiffuseAgent-MI improves accuracy by up to 5.1 points over prior self-evolving agents while more than doubling mutual-information faithfulness and human-interpretability agreement. Our analysis shows the energy term and the verifier are complementary: the former guarantees distributional faithfulness, the latter trajectory-level faithfulness, and only their combination closes both gaps.
Identifying the ``Inflamed'' immunophenotype in Gastric Adenocarcinoma predicts immunotherapy response but requires an expensive 10-gene RNA signature. While deep learning on standard H\&E slides offers a scalable alternative, conventional binary classifiers oversimplify continuous RNA data and introduce label noise. To resolve this, we propose VITA (VIrtual Transcriptomic Approximation). By aligning H\&E and RNA into a joint latent space during training, VITA requires only standard H\&E at inference to retrieve morphologically similar historical cases and approximate the continuous RNA signature. Achieving 0.72 classification accuracy and a 0.66 Spearman correlation, VITA provides a cost-effective ``virtual transcriptomics'' pre-screening tool that preserves the continuous phenotypic spectrum without requiring genomic sequencing.
Recent image-generation models and multimodal agents can produce high-quality visuals for increasingly complex visual communication tasks. Yet their raster outputs remain difficult to use directly because meaningful content and relationships are flattened into pixels, preventing users from inspecting, modifying, rearranging, or reusing individual components. We formulate image-to-editable reconstruction, which recovers a structured, directly manipulable artifact from a raster image while preserving its visual and semantic content. The central challenge is to jointly satisfy Fidelity and Editability, which often trade off in practice. To study this task, we introduce DrawAI, comprising an agentic benchmark, DrawAI-Bench, and a reconstruction workflow, DrawAI-Flow. DrawAI-Bench spans scientific figures, presentation slides, posters, and diagrams, combining real and AI-generated images to reflect practical visual-creation scenarios. It evaluates Fidelity and Editability through a hybrid protocol of 39 criteria: deterministic rule-based metrics measure properties with direct correspondences, while asset-specific vision-language rubrics capture semantic and perceptual qualities for which exact matching is misleading. Besides, we propose DrawAI-Flow, a two-stage agentic workflow in which a Parser Agent turns extracted elements evidence into an explicit reconstruction plan, and a Reconstruction Agent realizes the plan as executable graphics code through an iterative code-render-validate-revise loop. On DrawAI-Bench, we systematically evaluate thirteen models across five agent harnesses to study the effects of model capability, harness choice, and workflow design. The results show that reconstruction quality and costs vary substantially across model-harness configurations, while DrawAI-Flow consistently improves editable structure.
AI-generated image detectors deployed in open-world environments inevitably face distribution shifts as new and stronger generative models continue to emerge. Although existing methods improve cross-generator generalization through better representations or training data construction, they typically follow a static train-once-and-deploy paradigm and cannot adapt after deployment. In this work, we study open-set AIGC image detection from a test-time adaptation perspective. We propose Test-Time Curriculum (TTC), a simple and model-agnostic framework that adapts a detector on unlabeled test data through curriculum-based self-training. TTC starts from highly reliable pseudo-labeled samples and progressively incorporates harder yet informative cases, while enforcing class-balanced selection to reduce biased updates under generator shift. To further improve pseudo-label quality, we introduce Cross-Scale Pseudo-Label Refinement, which aggregates complementary evidence across multiple resolutions for more reliable adaptation, and applies noisy-or fusion at inference to strengthen final predictions. In addition, we construct AIGCGuard, a new benchmark containing 3,100 representative real images and 124,000 generated images from 40 of the most advanced open-source and proprietary text-to-image models. Extensive experiments on five benchmarks show that TTC substantially improves overall detection performance under diverse unseen-generator shifts, establishing a practical and effective test-time adaptation framework for open-set generated image detection.
Autoregressive image generators are commonly pretrained with token-level cross-entropy under teacher forcing, yet evaluated by the distributional quality of decoded images. This creates an objective mismatch, because categorical errors have unequal image-level consequences, and a context mismatch, because inference conditions on model-generated histories. We introduce FD-loss post-training, which adapts a pretrained discrete generator using representation-space Fr\'echet distance as the sole objective. A dual-pass scheme first constructs detached rollout contexts through gradient-free generation under the model's native inference configuration, then performs differentiable replay with a probability-level straight-through estimator (STE) that preserves hard argmax decoding in the forward pass while propagating image-level gradients through temperature-scaled probabilities. Only the generator is updated, while the tokenizer and feature extractors remain frozen. Across eight completed configurations from four generator families on class-conditional ImageNet at $256\times256$, FD-loss post-training reduces FID and $\mathrm{FD}_{r6}$ by 41.4% and 52.0% on average. The strongest FID result improves from 2.42 to 1.43 without adding parameters or inference steps.
Vision-language MoE batches contain different numbers of image and text tokens. Image resolution, image count, tiling, and prompt length all change this token mix. We call the standard token-level Switch auxiliary loss Std-Aux. Std-Aux balances only the mixed load, so large image and text load errors can cancel at one mix. On our main model, the same trained router shows more than a fivefold change in load imbalance across image resolutions. We hold the image and text load profiles fixed and derive the exact load curve as the token mix varies. The image-text load gap controls sensitivity to the token mix. Physical preprocessing can also change the conditional profiles. The fixed-profile law excludes such changes. To design a remedy, we examine the router input structure. Image and text occupy distinct regions, while visual tokens group strongly by source image. The modality boundary motivates separate image and text terms. The image boundary motivates one equal-weight routing instance per image. ReBA, or Relax Within, Balance Across, implements both choices. Across four split backbones, ReBA lowers load on every reported benchmark input while keeping mean task accuracy comparable to Std-Aux. ReBA also lowers average load over the tested range and worst physical load under resolution and tiling shifts. Code is available at https://github.com/ZiangWu-77/ReBA.
Recent advances in image generation and editing have made prompt quality a key bottleneck for e-commerce creatives. Vision-language models (VLMs) can generate image-editing prompts from product images and metadata, but further improving their prompt-writing capabilities requires post-training with feedback from the generated images. Group Relative Policy Optimization (GRPO) is a natural framework for such outcome-level reward optimization. However, it assigns credit only at the full-prompt level, even though image quality often depends on specific design elements such as composition, background, and the presentation of selling points. Existing fine-grained credit assignment methods typically require step-level supervision or learned critics. To address this, we propose EAGLE-GRPO (Element-Aware Group Learning for E-Commerce Image Generation), which decomposes the group-centered reward over predefined elements. We cast element-level credit assignment as a kernel ridge regression problem and derive a closed-form solution, without additional rollouts or separate credit-assignment models. This yields interpretable per-element advantages and more precise policy updates. Experiments show that EAGLE-GRPO sustains performance gains over more training steps before plateauing and generates prompts that produce higher-quality e-commerce images than competitive VLM prompt-writing baselines.
Representation Transfer of Foundation Models for Ultra-Widefield Retinal Imaging
基础模型在超广角视网膜成像中的表征迁移
Gong, Mingya Alexa, Ma, Da, Budimir, Lovre Antonio, Matovinovic, Ivana, Loncaric, Sven, Ju, Myeong Jin, Zhou, Yukun, Wagner, Siegfried K., Keane, Pearse A., Sarunic, Marinko V.
Abstract
Despite the widespread adoption of foundation models as feature extractors for medical imaging, relatively little is understood about how different pretraining strategies influence the transferability of learned representations to weakly supervised ophthalmic imaging tasks. We investigate this question in ultra-widefield (UWF) retinal imaging by evaluating foundation model representations within a patch-based multiple instance learning (MIL) framework for disease classification on UWF images. We compare Vision Transformer encoders pretrained with supervised, Masked Autoencoder (MAE), and self-distillation objectives, while keeping the downstream aggregation architecture unchanged. Within a controlled comparison of ViT-B encoders pretrained on ImageNet-1k, the choice of pretraining objective substantially influenced frozen representation transfer, with supervised and self-distillation-based models outperforming MAE. A contemporary DINOv3 model pretrained at a larger scale achieved the strongest overall performance, with a quadratic weighted kappa of 0.863 for five-class diabetic retinopathy grading, comparable with DINOv1. Attention analysis further revealed distinct patch-aggregation behaviours associated with the different pretrained representations, while partial fine-tuning substantially reduced the performance gap for MAE. These findings suggest that pretraining strategy influences both representation transferability and the subsequent aggregation of patch-level evidence within MIL, resulting in differences in downstream classification performance.
InstancePin: Instance-Addressable Layout-to-Image Diffusion via Coordinate Pinning
InstancePin：通过坐标固定实现实例可寻址的布局到图像扩散
Wu, Chaoyue, Zhang, Yunfei, Wu, Si
Abstract
Layout-to-image diffusion models have achieved impressive semantic controllability by conditioning generation on category-level segmentation maps. However, such category-aligned control is not necessarily instance-addressable: multiple nearby objects from the same category are often treated as a shared semantic region, leading to ambiguous boundaries, averaged appearances, and feature confusion among instances. This limitation is particularly evident in urban scene synthesis, where small and crowded pedestrians or vehicles require fine-grained instance separation while preserving global scene consistency. In this paper, we propose InstancePin, an instance-addressable layout-to-image diffusion framework that pins each object instance with an explicit coordinate anchor. Instead of directly injecting instance masks into the pretrained backbone, InstancePin introduces an independent instance-aware adapter to preserve the category-level generation prior while learning instance-specific spatial control. For each instance, its center coordinate is encoded with Fourier features and projected into a coordinate token, which serves as a spatial anchor queried by latent image features through coordinate pinning attention. To make these anchors spatially meaningful, we further supervise the coordinate attention maps with instance regions, encouraging each coordinate token to activate its corresponding object area. Finally, an instance-mask guided fusion module routes pretrained backbone features to non-instance regions and adapter features to instance regions, enabling local instance refinement without sacrificing global semantic fidelity. Extensive experiments on Cityscapes demonstrate that InstancePin mitigates instance entanglement in dense layouts and improves both image fidelity and semantic consistency.
Many physical attributes are \emph{irreversible}: ice melts but does not re-freeze, paper chars but does not un-burn. Do video generators respect this? We show the question is hard to measure, and that what can be measured reliably is \emph{development} rather than reversal. Metrics of local reversal are null-degenerate: a per-clip violation rate scores $0.50$ on pure noise, and a variance-normalized reversal residual sits at its noise ceiling. What survives null-testing is a two-part protocol: progress (a directional attribute correlation) and a stasis rate. Under this protocol, generated video separates cleanly from real footage, and the gap is human-validated. Across seven text-to-video models, real reference footage advances ($\rho{=}{+}0.40$, $35\%$ static) while every generator shows near-zero progress and $92$--$100\%$ stasis; nine annotators rate real footage far above generated ($2.75$ vs.\ $0.99$ on a $0$--$4$ scale). The reliable finding is \emph{under-development}: generators barely advance irreversible attributes rather than reversing them. As a complementary mechanism, we show that post-hoc readout guidance is gameable, whereas enforcing monotonicity by construction in a disentangled attribute latent removes the gameable readout, validated in controlled and semi-synthetic settings.
Where Does Generative Difficulty Reside? An Empirical Study of Target Representations
生成困难源于何处？目标表示的实证研究
Plocher, Marcel, Schölkopf, Bernhard, Geiger, Andreas, Gao, Gege
Abstract
The target representation defines the distribution an image generator must learn, yet it is often treated as an interchangeable interface. This assumption is particularly questionable for continuous masked generators, which combine contextual inference from visible tokens with conditional modeling of each missing token. We study raw pixels, SD-VAE latents and DINOv2 as well as MAE representation-autoencoder features within a unified masked autoregressive rectified-flow model. Under a shared ImageNet training budget, these spaces exhibit distinct optimization and inference regimes. DINOv2 converges fastest in both iterations and computation but benefits strongly from a wider local denoiser and direct context fusion. Pixels optimize substantially more slowly and require a different prediction, masking, and guidance configuration. MAE reconstructs images more faithfully and exhibits clear semantic clustering, yet produces generations substantially worse than DINOv2. The representations also respond differently to classifier-free guidance and occupy distinct precision-recall trade-offs. Together, our results show that compression, reconstruction fidelity, token dimensionality, and visible semantic clustering do not individually predict generative behavior. Instead, target representations redistribute difficulty across contextual modeling, per-token denoising, and inference-time distributional control.
Chinese Translation
目标表示定义了图像生成器必须学习的分布，但它通常被视为可互换的接口。对于连续掩码生成器，这一假设尤其值得怀疑，因为它结合了来自可见标记的上下文推理与对每个缺失标记的条件建模。我们在统一的掩码自回归修正流模型中研究了原始像素、SD-VAE 潜变量、DINOv2 以及 MAE 表示自编码器特征。在共享的 ImageNet 训练预算下，这些空间表现出不同的优化和推理机制。DINOv2 在迭代和计算上收敛最快，但在更广泛的局部去噪器和直接上下文融合方面受益显著。像素的优化速度明显较慢，并且需要不同的预测、掩码和引导配置。MAE 更忠实地重建图像，并展现出明显的语义聚类，但生成的结果明显逊色于 DINOv2。这些表示对无分类器引导的响应也不同，并占据了不同的精度-召回权衡。综合来看，我们的结果表明，压缩、重建保真度、标记维度和可见语义聚类并不能单独预测生成行为。相反，目标表示在上下文建模、逐标记去噪和推理时的分布控制之间重新分配了困难。
WiFuse: An Attention Mechanism for Human Activity Recognition using Fused CSI Amplitude and Delay-Doppler Channel Features
WiFuse：一种基于融合CSI幅度和延迟-多普勒通道特征的人体活动识别注意力机制
Fernandes, Alison M., Del Monego, Hermes I., Chang, Bruno S., Munaretto, Anelise, Fontes, Hélder M., Campos, Rui L.
Abstract
Recently, Wi-Fi sensing has played a significant role in Human Activity Recognition (HAR), as it enables the detection of various activities using only Wi-Fi signals, ensuring privacy and remaining non-intrusive for the user. However, environmental characteristics such as reflective surfaces, hardware offsets, and other physical impairments affect recognition by the neural network, subsequently causing errors and significantly reducing model accuracy. To overcome this problem we present the WiFuse framework, a dual-stream Channel State Information (CSI) framework for human activity recognition (HAR) that pairs denoised time-domain amplitude variations with 2D-FFT-derived Delay-Doppler motion representations computed from the sanitized channel phase. The fused representation feeds a hybrid ResNet-Temporal Convolutional Network (TCN) neural architecture augmented with channel and spatio-temporal attention, where the ResNet extracts spatial-spectral features and the TCN models long-range temporal dependencies; a decoupled two-stage transfer learning strategy is employed to improve optimization stability and feature reuse. We conduct extensive experiments on two public datasets, including comparisons against state-of-the-art methods and alternative hybrid architectures, ablation studies, and cross-dataset and domain-adaptation evaluations. The proposed framework reaches an overall accuracy of up to 95.28% across the four environments of the XRF55 dataset and up to 98.20% on the multi-user Wi-MIR dataset. Overall, the results indicate that combining amplitude and Delay-Doppler representations within a dual-stream strategy, enhanced by transfer learning, improves recognition performance under conditions that typically degrade deep neural networks, such as class overlap, multipath propagation, noise, and interference.
PixelSR: Efficient Screen Content Super-Resolution via Pixel Classification
PixelSR：通过像素分类实现高效的屏幕内容超分辨率
Li, Zhiheng, Chen, Lei, Zhou, Jie, Lu, Jiwen
Abstract
Screen content images are generally composed of texts and graphics. Compared to natural images, these man-made images contain a large quantity of sharp but repetitive structures. However, existing works in screen content super-resolution underutilize the special characteristics of screen content, leaving a large room to improve model performance and speed up. In this paper, we propose PixelSR, a simple yet effective method to improve super-resolution performance but with faster inference speed. To improve model performance, we classify pixels via pixel binning to compute content attention in the training phase. Specifically, after binning pixels into content-dependent groups, content attention is aggregated from pixel features within each group to introduce a content-dependent and non-local receptive field for every pixel. In the testing phase, we utilize the properties of self-repetitiveness and redundancy in screen content to speed up inference without the loss of model performance. We divide targeted high-resolution pixels into three types, which are unique pixels, repeated pixels, and background pixels for each test image. We conduct conventional network processing on unique pixels and cache their predictions in the on-the-fly lookup table. For repeated pixels which have appeared in unique pixels, we directly retrieve prediction results from the lookup table without network processing. For background pixels, we use the nearest neighbor algorithm to generate high-resolution pixels. The on-the-fly lookup table is cleaned and repeats the procedure above for the next test image. Experiments show our PixelSR achieves state-of-the-art performance with shorter inference time in screen content super-resolution.
Audio-driven emotional talking face generation aims to synthesize realistic videos with expressive facial dynamics. However, existing methods struggle to balance controllability and visual fidelity. Although implicit representations capture rich semantics, they lack structural guidance, often resulting in averaged emotional expressions. In contrast, explicit geometric methods offer better control over facial expressions but tend to sacrifice high-frequency texture details. To address it, we propose GemTalk, a diffusion-based framework that combines the semantic richness of implicit representations with the structural precision of explicit geometric priors. We introduce a Vision-guided Audio Emotion Projection (V-AEP) module to extract implicit emotional lip and expression features. At the same time, a Diffusion-based Geometric Priors Generator (D-GPG) generates identity-aware blendshape coefficients as explicit structural priors. Crucially, our Geometry-guided Emotion Modulation (GEM) module leverages these geometric priors to recalibrate the magnitude of implicit features, enabling precise, continuous control over emotional expressions, especially emotion intensity, without sacrificing visual quality. Extensive experiments show GemTalk achieves superior performance in photo-realism, and facial emotional dynamics.
CopyCat: Improving Fine-Grained Subject Consistency in Subject-to-Image Models within Seconds
CopyCat：在几秒钟内提高图像生成模型中的细粒度主题一致性
Zheng, Peng, Liu, Ruiqi, Ma, Rui, Wu, Zuxuan
Abstract
Recent subject-to-image models have achieved impressive progress in personalized image generation, yet they still struggle to preserve fine-grained subject-specific details. A major reason is the lack of high-quality fine-grained identity supervision: real paired data are expensive to collect, while synthesized training pairs often preserve only coarse subject appearance and fail to capture subtle subject-specific details. In this work, we propose CopyCat, a lightweight model-refinement framework that improves fine-grained subject consistency within only a few seconds. CopyCat performs a one-time refinement of a pretrained subject-to-image model by attaching a lightweight Fine-grained Consistency LoRA (FCLoRA) and optimizing it using a single proxy image, which is used as both the conditioning image and the reconstruction target. This exact self-reconstruction objective substantially simplifies the optimization task, enabling effective fine-grained refinement within only a few seconds. The refinement is performed only once; the resulting model can be directly applied to diverse unseen reference subjects and prompts without further subject-specific optimization. We further revisit subject-to-image LoRA training in double-stream diffusion transformers and find that adapting only the visual stream consistently improves subject consistency. Extensive experiments on DreamBench and XVerseBench demonstrate consistent improvements in fine-grained subject consistency across representative subject-to-image models under both single- and multi-subject settings.
Breaking the Horizontal Prior: From Long-Tailed Orientation Bias to Roll-Robust Monocular Depth Estimation
打破水平先验：从长尾导向偏差到稳健的单目深度估计
Tang, Kaihua, Xia, Ziqing, Zheng, Xiaoxu, Zhang, Xiaoxue, Mi, Michael Bi, Xu, Zhan, Chen, Dave Zhenyu
Abstract
Despite recent advances in Monocular Depth Estimation, state-of-the-art depth foundation models remain vulnerable to robustness issues. Particularly, even slight camera rolls can result in substantial degradation in depth estimations. We attribute this problem to a previously overlooked phenomenon, termed the Horizontal Prior, which is a manifestation of long-tailed distribution bias: most training images are captured in approximately horizontal orientations due to human visual preferences and photographic habits. While intuitive remedies such as re-balanced data augmentation and horizon leveling provide partial improvements, they fail to fully address the issue. In this paper, we introduce Invariant Depth Constraint (ID-Constraint), a training-time supervision strategy that improves roll robustness by fine-tuning and jointly regularizing the depth backbone with a series of geometric and spatial reasoning tasks. These auxiliary objectives encourage the backbone to learn rotation-stable, depth-relevant representations, while the auxiliary prediction heads are discarded after training, leaving the original inference architecture unchanged. Extensive experiments on five benchmark datasets across four roll settings demonstrate the effectiveness of the proposed method.
BRIC-Net: Boundary-Reliable Illumination-Color Interaction for Remote Sensing Image Deshadowing
BRIC-Net：边界可靠的照明-颜色交互用于遥感图像去阴影
Lu, Wei, Liu, Yi, Si-Bao
Abstract
Shadows in remote sensing images obscure surface appearance and disrupt radiometric continuity, reducing the reliability of visual interpretation and downstream analysis. Remote sensing image deshadowing is an ill-posed inverse problem that requires spatially varying illumination recovery while preserving chromatic and radiometric consistency in non-shadow regions. Existing methods commonly rely on hard shadow masks for compensation or directly regress RGB intensities. Hard masks may inadequately model gradual penumbra variations and are sensitive to localization errors, often producing residual shadows or halo artifacts; direct RGB regression entangles illumination recovery with chromatic reconstruction and can introduce color casts. To this end, we propose the Boundary-Reliable Illumination-Color Interaction Network (BRIC-Net), which decouples these failures at different representation levels. A Lightness Reliability Prior (LRP) derives reliability-aware guidance from CIELAB statistics. Boundary-Adaptive Gated Mixing (BAGM) performs gated interpolation between shallow RGB and lightness features around uncertain transitions, while Spatial-Channel Mutual Modulation (SCMM) coordinates deeper spatial and channel responses for appearance-preserving illumination recovery. BRIC-Net achieves 29.46~dB full-image peak signal-to-noise ratio (PSNR) on AeroDS-Syn and 27.96~dB on SRGTA. It also obtains the lowest Perception-based Image Quality Evaluator (PIQE) scores on AISD and AeroDS-Real. Region-wise evaluations and component ablations further support its effectiveness in shadow recovery and non-shadow preservation.
Proteus: A Truncation-Robust Entropy Model for Progressive LiDAR Compression
Proteus：一种针对渐进式LiDAR压缩的截断鲁棒熵模型
Qiu, Yihan, Lin, Xiaodong, Zhao, Baoquan, Jiao, Hailong, Li, Ge
Abstract
LiDAR point clouds provide explicit, deterministic physical boundaries critical for collaborative safety-critical perception. However, wireless channels inherently impair and corrupt transmitted signals. Existing robust frameworks (such as deep JSCC or MDC) attempt to counter these channel impairments through statistical or parametric estimation, turning exact physical measurements into unverified algorithmic estimates. To address this, we propose Proteus, a learned LiDAR codec operating on 2D range images. By decoupling the frame representation into independent coders for the \textbf{sig}nificant range bit-planes (SIG) and the \textbf{ins}ignificant range bit-planes and attributes (INS), Proteus achieves overall stream-level truncation robustness. The non-truncatable SIG block encodes the most significant range bit-planes to establish a necessary, self-contained perceptual lower bound, below which the reconstructed point cloud is severely degraded. Meanwhile, INS employs bit-plane slicing representation and coding, ensuring that range truncation mathematically maps to a deterministic spatial precision degradation. Subordinate attributes are reconstructed via a hybrid lossless-predictive method, leveraging the decoded geometry as a strong structural prior for fine-grained approximation. Furthermore, strategic ordering within INS prioritizes geometry over attributes under bandwidth drops. Experimental results on the Waymo Open Dataset and SemanticKITTI demonstrate that Proteus tolerates up to approximately 70\% bitstream truncation, while outperforming established standards (G-PCC, Draco, and JPEG XL) and the representative learned compressor Unicorn under ideal channel conditions.
Event cameras offer microsecond-level temporal resolution and high dynamic range, potentially facilitating motion-blur-free panoramic imaging from fast rotational scanning. Nonetheless, existing optimization-based methods remain computationally demanding, while prior learning-based reconstruction methods are largely designed for perspective imagery and lack geometry-aware support for panoramic outputs. We present E2Pano, a geometry-guided event-to-panorama pipeline with an end-to-end learnable photometric reconstruction stage. Our framework preserves real spherical coordinates from geometric mapping throughout the pipeline, employs a lightweight enhancement module with frequency-domain supervision to bridge the event-image domain gap, and leverages a spherical Transformer with 3D positional embeddings for photometric reconstruction. Experiments on synthetic data and captured rotational scans show improved reconstruction quality and lower photometric reconstruction cost than optimization-based baselines, together with encouraging transfer to real captures under our acquisition protocol despite training purely on synthetic data. Additionally, we construct PanoScan, a dataset with 4,370 synthetic and 30 real-world panoramic scenes paired with event streams. Our dataset and code will be released.
Zero-shot anomaly detection (ZSAD) aims to detect anomalies and localize defective regions in unseen target domains without target training data. Recent ZSAD methods build on pretrained vision models, particularly CLIP, and construct normal and anomaly references from textual prompts or learnable visual representations. These methods perform anomaly discrimination primarily in spatial feature spaces, where subtle changes in texture, boundaries, and local structures can be confused with normal appearance variations. Although inconspicuous spatially, such defects can disrupt local texture regularity or boundary continuity, inducing response deviations across frequency bands. However, existing ZSAD methods do not explicitly model these frequency-dependent characteristics. Our image-domain analysis reveals that local defects exhibit spatial-frequency deviations from normal references across low-, middle-, and high-frequency bands, indicating that anomaly evidence is not universally dominated by high-frequency responses. Motivated by this observation, we propose FreqAnchorAD, a frequency-aware framework that organizes frequency-enhanced responses for anchor-relative anomaly discrimination. Specifically, the Local Frequency Compensation Module (LFCM) enhances intermediate patch tokens with local spatial-frequency cues. The Frequency-Deviation Anchor Projector (FDAP), our core discrimination module, organizes enhanced responses along a source-derived channel coordinate and measures anomaly evidence through relative similarity to normal and anomaly anchors. Finally, Asymmetric Anchor Supervision (AAS) stabilizes normal-anchor alignment while preserving diverse anomaly patterns. Experiments on thirteen industrial and medical benchmarks show that FreqAnchorAD achieves state-of-the-art mean performance in image-level anomaly recognition and pixel-level defect localization.
AeroLLE: Constrained Pseudo-Supervision for Nighttime Aerial Image Enhancement with the AeroNight-1.5K Benchmark
AeroLLE：基于约束伪监督的夜间航空图像增强及AeroNight-1.5K基准
Lu, Wei, Liu, Hongyuan, Chen, Si-Bao
Abstract
Nighttime aerial image enhancement is challenged by spatially nonuniform exposure, mixed illumination, and weak structural evidence, while registered normal-light targets are difficult to capture from moving platforms. Generated normal-light images provide practical appearance guidance but may alter geometry or texture. We introduce \aeronight{}, comprising 1,500 real nighttime aerial RGB images: 1,300 inputs are associated with manually screened pseudo-references, and 200 inputs support unpaired evaluation. We propose AeroLLE, a two-stage framework that first recovers visibility with an HVI Base Enhancer and then performs Spatially Adaptive Exposure--Color Calibration (SAECC). After the Base Enhancer is selected and frozen, SAECC predicts bounded, low-resolution RGB gain and bias fields, restricting the magnitude and spatial variation of the second-stage correction. Experiments under complementary pseudo-paired and unpaired protocols demonstrate improved agreement with screened appearance targets, together with more balanced exposure and color correction across diverse nighttime aerial scenes. These results support constrained, stage-specific calibration as a practical strategy for learning from generated appearance guidance when registered aerial references are unavailable.
Recent advances in large vision-language models (LVLMs) have enabled long-video understanding and analysis. However, processing the large number of frames in a video incurs substantial computational overhead. Existing methods reduce LVLM inference costs by scoring frame-query relevance before inference and selecting keyframes accordingly. Nevertheless, the distribution of relevant frames varies across queries, and these methods often need to score hundreds or thousands of frames. To address this limitation, we propose CSES, a training-free semantic keyframe selector that adaptively determines the numbers of frames to score and keyframes to select. CSES estimates the prominence of the frame-query relevance profile to guide active acquisition and adapt the temporal coverage of each input. It then formulates keyframe selection as a coverage problem that jointly accounts for semantic relevance, temporal redundancy, and visual redundancy. Active acquisition and keyframe selection terminate based on coverage saturation. The selection objective is monotone and submodular, enabling greedy optimization with a standard approximation guarantee. Experiments with four LVLMs on two benchmarks show that our method preserves accuracy while scoring $4$-$13\times$ fewer frames and selecting $18.4\%$-$20.5\%$ fewer input keyframes than existing baselines. CSES further achieves a $3.1$-$5.4\times$ speedup in frame selection over baselines.
Robust detection of generated images is critical to counter the misuse of generative models. Existing methods primarily depend on learning from human-annotated training datasets, limiting their generalization to unseen distributions. In contrast, large-scale vision models (LVMs) pre-trained on web-scale datasets exhibit exceptional generalization power through exposure to diverse distributions, offering a transformative paradigm for this task. However, our experimental results reveal that LVMs pre-trained on natural-image-dominated data can effectively capture the features of both natural and generated images, yielding comparably low losses and thus limited discriminative capacity between them. This prompts a key question: When and how do LVMs exhibit different behaviors when capturing features of natural and generated images? This investigation reveals an insight: during unlearning, LVMs exhibit disparate forgetting dynamics with feature degradation for generated images escalating faster than natural ones. Inspired by the disparate dynamics, we introduce two detection methods: 1) data-free detection, which prunes model parameters to induce unlearning without data access, and 2) data-driven detection, which optimizes LVMs to unlearn knowledge tied to generated images. Extensive experiments conducted on various benchmarks demonstrate that our unlearning-based approach outperforms conventional detection methods. By recasting the detection task as a problem of machine unlearning, our work establishes a new paradigm for generated image detection.
Frozen vision foundation models are commonly evaluated through a single global image embedding, but this interface can conflate missing information with information lost at readout time. We study this distinction by keeping a pretrained vision encoder frozen and varying only the readout applied to its final patch tokens. We compare standard global readouts against a lightweight foveated readout, which attention-pools patch tokens using a learned or question-conditioned query, and against an oracle readout with access to the annotated target region. We evaluate these interfaces on three localized binding problems: a controlled synthetic color--shape binding task under clutter, a color-free crowded shape-detection variant, and a GQA-derived natural-image task where paired questions ask for the colors of different same-category objects in the same image. Global readouts perform near perfectly when the synthetic target appears alone, but collapse under clutter and counterfactual target edits, whereas the foveated readout recovers most of the oracle-accessible signal. On the GQA-derived task, question-independent global image vectors improve only modestly over question-only priors, while question-conditioned foveation substantially improves paired localized color accuracy. A counterfactual nuisance-to-signal ratio explains the synthetic failures: global pooling dilutes localized label-changing evidence while exposing the probe to nuisance variation from irrelevant objects. These results indicate that apparent spatial blindness in frozen vision models can arise from the global embedding interface rather than from an absence of spatial information in the frozen patch tokens.
MDTD-ArtIR: Benchmarking Image Editing and Restoration Models for Art Image Restoration under Texture-Overlay Degradations
MDTD-ArtIR：在纹理叠加退化下对艺术图像修复模型的基准测试
Vijendran, Mridula, Chen, Shuang, Shum, Hubert P. H.
Abstract
Restoring severely degraded visual media still remains a formidable challenge, as existing methods often hallucinate unnatural textures and contents, struggle with preserving color and texture, or fail to leverage partially retained image information. Existing restoration benchmarks assume known degradation operators and fail to capture the complex characteristics of artistic damage such as cracks, stains, and color/texture deviation. We introduce a controlled benchmark for blind restoration of semantic, semi-transparent image media degradations, accompanied by a new, publicly open degradation alpha texture mask dataset MDTD-Art. We present a new dataset and benchmark evaluating state-of-the-art universal restoration models against image editing and vision-language models across varying mask opacity levels. Our experiments demonstrate that image editing models consistently outperform specialized restoration architectures for arbitrary degradations, with performance gains amplified by structured prompt engineering emphasizing detail preservation and structural consistency. These findings position recoverable semantic information and prompt controllability as critical factors in art image restoration.
Multimodal large language models have advanced visual understanding, yet perception-intensive reasoning remains challenging. Recent latent visual reasoning methods introduce hidden-space computation before answering, but they often rely on costly intermediate supervision, such as bounding boxes, sketches, or interleaved rationales. These strategies focus on how latent states should be shaped, but do not explicitly assess whether the latent is useful for the final answer. We propose LUT, a latent reasoning framework trained with only standard VQA pairs. LUT centers training on Latent Utility at two levels. At the trajectory level, we propose Utility-Aware Latent Distillation SFT, which explores answer-relevant latent trajectories, selects qualified trajectories by their information gain, and distills more reliable and learnable supervision through curriculum learning. At the step level, we propose Latent Attribution Policy Optimization, which uses answer-to-latent attribution to differentially optimize latent steps during reinforcement learning. Experiments on perception-intensive visual reasoning benchmarks show that LUT outperforms previous latent reasoning methods and remains competitive with latent-text interleaved methods with lower annotation cost.
NISF++: Geometrically-grounded implicit representations of 3D+time cardiac function from 2D short- and long-axis MR views
NISF++：基于几何的隐式表示从二维短轴和长轴磁共振视图获取三维+时间心脏功能
Stolt-Ansó, Nil, Dannecker, Maik, Jia, Steven, McGinnis, Julian, Rueckert, Daniel
Abstract
Clinical acquisition in cardiac magnetic resonance (CMR) imaging involves obtaining cross-sectional planes of the heart along the radial and longitudinal directions. Despite these planes being 2D cross-sectional images of the heart, radiologists understand the 3D spatial and continuous temporal nature of the organ being imaged. The same can not be said about the conventional deep learning architectures used to process CMR images, which rely on in-plane and grid-based operations, and are hence unable to organically integrate information from all imaging planes. This paper builds upon previous work on neural implicit segmentation functions (NISF) to overcome unaddressed challenges in cardiac function modeling in the CMR domain. For a given subject, our architecture builds a shared 3D+time representations from all available acquisition planes regardless of orientation. By design, predictions along any imaging plane orientation are cross-sections of the same 3D representation, leading to spatio-temporal consistency across all slices. Moreover, our architecture makes the rotation and translation parameters of imaging planes learnable, allowing us to correct for the commonplace respiratory and patient motion between slice acquisitions under a rigid assumption. Furthermore, interpolation of intensities and segmentation can be performed in 4D at any desired resolution. We perform our study on a 120 subject sub-cohort of CMR imaging data from the UK-Biobank. Our in-plane segmentation performance is on-par with existing CMR segmentation methods and explore how the majority of failure cases arise from limitations in the ground-truth segmentation, for which our representations make predictions with better anatomical accuracy than its original training data. We also evaluate our motion-correction capabilities, displaying quantitative and qualitative improvements in slice alignment.
ChordVideo: One-Step, Training-Free, Temporally Consistent Video Editing via Low-Energy Transport
ChordVideo：通过低能量传输实现一步法、无训练、时间一致的视频编辑
Lao, Zhiqiang
Abstract
One-step text-to-image models enable training-free, inversion-free editing with only 1--2 network function evaluations (NFE), while ChordEdit stabilizes such edits through low-energy smoothing along sampling time. Applied independently to video frames, however, it produces temporal flicker and edit-strength drift. We introduce \textbf{ChordVideo}, which extends the same low-energy principle to video time through shared noise, motion-aligned causal aggregation of per-frame Chord fields, and an optional temporally smoothed proximal correction. We derive a warping-error bound that separates motion bias from stochastic flicker and predicts diminishing returns with larger temporal windows. On TGVE/DAVIS with two one-step backbones, ChordVideo reduces warping error by \textbf{78\%} and flicker by \textbf{49\%}, improves CLIP frame consistency by \textbf{9--10 points}, and increases background PSNR by about \textbf{1.5,dB}, while retaining \textbf{2 NFE/frame}. Compared with seven multi-step editors, it achieves competitive temporal consistency and source preservation using \textbf{10--60$\times$ fewer model steps per clip
Computer-Aided Design (CAD) underpins modern engineering, yet converting existing shapes into editable models still demands substantial expert effort. Most AI systems emit the entire CAD program in a single pass, never inspecting the intermediate geometry. In contrast, human engineers build a part feature by feature, checking after each operation what remains to be modeled. We introduce CADENA (Spanish for "chain"), a model that reconstructs a 3D mesh as a parametric CAD program, growing its sequence of operations one at a time and comparing the target with the currently predicted geometry at every step. We also address the lack of benchmarks for evaluating reverse-engineering methods on mechanical parts, introducing CADENA-Bench, a benchmark that measures performance across categories of mechanical parts. CADENA outperforms prior methods on CADENA-Bench and on the DeepCAD, Fusion 360, and MCB datasets. Code is available at https://github.com/zhemdi/cadena, model weights at https://huggingface.co/kulibinai/cadena, and CADENA-Bench at https://huggingface.co/datasets/kulibinai/cadena-bench.
Injection molding is the cornerstone of mass-producing plastic components. While current algorithms can automate mold design for basic geometries using standard two-piece molds, complex parts featuring undercuts, side holes, or re-entrant features present a significant challenge. These geometries often necessitate auxiliary components beyond the primary upper and lower molds. In practice, designing these intricate assemblies is a laborious process that relies heavily on expert knowledge. Furthermore, the scarcity of public datasets has hindered the development of effective learning-based solutions. To bridge these gaps, we introduce MoldCAD, a curated dataset that pairs complex single-body CAD parts with industry-standard mold assemblies. Each entry includes the upper and lower molds, parting surfaces, demolding orientations, and necessary auxiliary components. The dataset comprises 4,934 CAD models and over 3,850 mold assemblies, totaling more than 23k individual models. Building upon this dataset, we propose a comprehensive pipeline that predicts demolding orientations, identifies auxiliary components, and constructs parting surfaces to derive a complete, manufacturing-ready mold assembly for downstream CAD/CAM workflows. Our results demonstrate a promising path toward fully automated industrial mold design and contribute to the broader advancement of manufacturing-aware CAD generation.
Most current visual trackers adopt a matching-based architecture trained exclusively on tracking datasets, whose performance gains depend heavily on the length of the input context, and have now reached a bottleneck. While high-performance tracking increasingly relies on foundation models, existing methods use them monolithically, adapting a foundation model into a tracker or modify a segmentation foundation model into a tracking pipeline, which fails to exploit complementary strengths. Matching-based trackers excel at instance-level correspondence but lack semantic discrimination and fine-grained foreground perception, whereas segmentation foundation models produce precise masks yet struggle with instance discrimination and multimodal extension. Both paradigms also lack error-correction capabilities for long-term tracking. To address these issues, we propose ACTrack, an agentic coordination framework that treats heterogeneous models as invocable tools under an event-triggered mechanism. ACTrack coordinates a Tracker-based Instance Matching Tool for target discrimination, a SAM3 Motion Tool for mask-derived motion priors, a SAM3 Perception Tool for detecting distractors and instance-conflict cues, and a VLM Reprompt Tool activated only under persistent conflict to mitigate error accumulation. We design a complete tool-invocation trigger mechanism and an inter-tool coordination mechanism, enabling the complementary strengths of different model tools to be fully integrated. Experiments show that ACTrack substantially surpasses the strongest and the largest trackers on eight RGB benchmarks. Furthermore, a parameter-efficient adaptation strategy enables parameter sharing and reuse across tools, achieving unified multimodal tracking with only 30\% trainable parameters while substantially outperforming prior methods on multimodal benchmarks such as LasHeR, VisEvent, TNL2K, and DepthTrack.
MIDAL: Math Image Descriptions for Accessible Learning
MIDAL：可访问学习的数学图像描述
Popek, Rebeka, Ojha, Vaghawan, You, Young Hwan
Abstract
Many open educational resources are lacking in accessibility, especially in-depth image descriptions. In subjects like Science and Mathematics, however, it can be particularly difficult to write image descriptions since there can be many complicated expressions and names depending upon the course level. To help fill that gap in a small way, we introduce Math Image Descriptions for Accessible Learning (MIDAL), a math image-description dataset of 2,020 mathematical images spanning multiple educational levels, to aid in training vision language models to create image descriptions following accessibility best practices. We hope MIDAL is a valuable resource in enhancing the conversation and innovation regarding accessibility of STEM content in higher education. This dataset is however not just limited in math description generation but can also be used to fine-tune language models that can have improved mathematical reasoning and answers.
PhenoStitch: Training-Free Panoptic Crop Mapping from Satellite Image Time Series
PhenoStitch：无训练的全景作物映射方法基于卫星图像时间序列
Li, Xuechen
Abstract
Panoptic crop mapping requires both delineating individual agricultural parcels and assigning a crop type to each parcel from satellite image time series. Existing approaches typically rely on dense parcel-level annotations and task-specific model training, which limits their applicability to new regions and growing seasons. We introduce PhenoStitch, a panoptic crop-mapping pipeline that requires no task-specific gradient-based training. A frozen Segment Anything model first oversegments each patch into class-agnostic regions. For each region, optical NDVI and Sentinel-1 backscatter series are summarized by an analytic double-harmonic phenological signature. Adjacent regions are then merged into parcels by minimizing a Potts graph energy, and each parcel is classified by nearest-prototype matching using only (k) labeled parcels per class. A final topology-closure step produces the panoptic map. Under a matched budget of (k=20) parcels per class, corresponding to less than 1% of the available labels, PhenoStitch achieves 20.0 crop mIoU, 76.2 segmentation quality, and 6.2 panoptic quality on PASTIS-R under a 5-fold, 3-seed evaluation. It outperforms the evaluated frozen foundation-model, few-shot, and matched-budget supervised baselines under the same protocol, with a consistent ranking also observed on ZueriCrop. Ablation studies show that radar observations contribute the largest performance gain, while the graph-energy merge and compact phenological signature provide further improvements. These results demonstrate the effectiveness of combining label-free parcel delineation with few-shot phenological recognition for panoptic crop mapping under limited supervision.
Robust to intensity inhomogeneity, the local Chan--Vese (LCV) model extends the classical Chan--Vese (CV) image segmentation method by incorporating local statistical information around each pixel. Originally, the LCV model was solved using a finite difference scheme, following the approach used for the CV model. As an alternative to the finite difference scheme, a more efficient algorithm based on the Merriman-Bence-Osher (MBO) scheme was later developed for the CV model. In this paper, we derive a similar MBO-based algorithm to solve the LCV model and propose an efficient implementation. The algorithm is developed for both two-phase and multiphase segmentation, and an extension to color images is also discussed. To demonstrate the effectiveness of the proposed approach, we apply it to a variety of grayscale and color images, including medical and microscopy images.
PeCA: Palette Context Assisted Inference for Test-Time Paint-Bucket Colourisation on Animation Videos
PeCA：用于动画视频测试时油漆桶着色的调色板上下文辅助推理
Lin, Dongheng, Jiao, Jianbo
Abstract
In animation production, paint-bucket colourisation for hand-drawn animation is a labour-intensive procedure that assigns each enclosed region in line sketches a colour from reference design sheets. Recent automatic paint-bucket colourisation pipelines mirror this workflow via region correspondence, but correspondences can be brittle when regions are ambiguous fragments without proper context. In this paper, we propose Palette Context Assisted (PeCA), a new training-free, plug-and-play framework for animation video colourisation that aims to close this gap at test-time via reasoning over spatial and temporal contexts. Extensive experiments on existing benchmarks and a newly introduced long-video test case show consistent performance boosts.
Monocular panoramic SLAM benefits from substantial visual overlap under large camera rotations, yet remains prone to errors caused by camera tilt, scale drift, and false loop closures. We show that a frozen panoramic geometry foundation model provides useful internal cues beyond its explicit geometric outputs: intermediate tokens encode gravity in the camera frame, while cross-view attention provides a compatibility cue for potential revisits. Building on these cues, we present HALO-SLAM. A gravity readout enables IMU-free spherical upright canonicalization. For loop closure, we introduce a cost-aware three-stage cascade combining DBoW2 event-level retrieval, attention-based compatibility filtering, and dense geometric validation through symmetric submap augmentation. Accepted revisits yield pixel-aligned 3D--3D correspondences in both local gauges, from which robust $\mathrm{Sim}(3)$ constraints are estimated and jointly optimized with sequential constraints in a global pose graph. Across 125 sequences from five real-world panoramic benchmarks, our method achieves \textbf{100\%} sequence success (\textbf{125/125}) under the stated criterion and the lowest ATE among the evaluated methods on all five benchmarks, reducing ATE by \textbf{30--88\%} relative to the best ERP-native baseline on each benchmark.
Swimm3R: Splatting with Medium-aware SfM for Underwater 3D Reconstruction
Swimm3R：一种结合介质感知结构光束（SfM）与水下贝塔点云的水下三维重建框架
Kweon, Minseong, Sattar, Junaed
Abstract
We propose Swimm3R, a unified framework that combines medium-aware structure-from-motion (SfM) with Underwater Beta Splatting to address scattering- and attenuation-induced failures in underwater 3D reconstruction. Swimm3R distills in-air geometric priors into a feed-forward backbone and uses a physics head to regress underwater image-formation parameters, camera poses, and restored point clouds. Additionally, we introduce Underwater Beta Splatting, which extends Gaussian splatting with Beta primitives and scattering-aware geometric gradients for stable underwater geometry representation. We further establish the Barbados underwater video dataset to demonstrate the effectiveness of our method in challenging underwater environments. On this dataset, Swimm3R robustly recovers underwater scene structure under challenging scattering conditions, yielding coherent seafloor geometry. Using these predicted point clouds, the proposed Underwater Beta Splatting improves average PSNR by $1.47$ dB over WaterSplatting while increasing downstream localization performance by $2.0$ and $2.4$ percentage points in RRA@15 and RTA@15, respectively.
Large vision--language models (VLMs) can reason step by step about complex visual scenes, but this open-ended, autoregressive chain-of-thought (CoT) approach is poorly suited to safety-critical, rule-governed settings such as industrial surveillance, where decisions must be bounded, deterministic, and low-latency. Because CoT inference cost scales jointly with reasoning length and the number of concurrent streams, it creates a throughput bottleneck that precludes the real-time, multistream monitoring required for industrial accountability. Here we present MonitorVLM-v2, a deployment-oriented framework that recasts VLM-based safety assessment as probabilistic inference over a finite regulatory decision space, compressing multimodal reasoning into single-step rule-ID predictions and reducing decoding from a variable-length sequence to a single token. We introduce symbolic policy optimization (SymPO), a novel contrastive policy optimization algorithm that sharpens decision boundaries within this finite symbolic space, together with an entropy-driven triage mechanism that routes uncertain predictions to human reviewers for expert confirmation. In a four-month prospective deployment across 10 concurrent camera feeds in an operational underground mining facility, MonitorVLM-v2 achieved a 19.45-fold increase in inference speed and identified 2.78 times as many confirmed violations as the site's routine manual inspection workflow, demonstrating the practical value of compressed symbolic decision-making for real-time, auditable industrial monitoring.
Location-Aware Fine-Grained Representation Learning for Medical Vision Foundation Models
基于位置感知的细粒度表示学习用于医学视觉基础模型
Kang, Myeongkyun, Yang, Yanting, Li, Xiaoxiao
Abstract
Fine-grained visual representations are essential for medical image analysis, particularly when diagnostically relevant evidence is subtle and spatially localized. Modern transformer-based medical vision encoders must therefore learn patch-level representations that are both clinically meaningful and spatially consistent. Without these properties, large vision-language models (LVLMs) operate on an ambiguous visual foundation, limiting their ability to generate clinically reliable and spatially grounded responses. However, existing training strategies for medical vision encoders rarely achieve both objectives. Image-text alignment provides clinically meaningful supervision primarily at the image level, leaving the spatial localization of diagnostic evidence weakly constrained. In contrast, self-supervised learning promotes spatial consistency but lacks the semantic supervision needed to distinguish visually similar yet clinically distinct regions. To address this gap, we present LoFi, a medical vision foundation model built on location-aware fine-grained representation learning. LoFi trains a vision encoder with a lightweight large language model under grounding and grounded captioning objectives. Because these objectives require predicting location from clinical text and vice versa, spatial consistency emerges without any explicit patch-level regularization. To enable training at scale, we construct MedG, a large-scale medical grounding dataset of 4.48M image-text-box triplets curated from 84 datasets spanning 7 modalities. Across phrase grounding, visual question answering, and region-based organ classification under perturbations, LoFi consistently outperforms general-purpose and medical vision foundation models as well as state-of-the-art LVLMs. Code is available at https://github.com/myeongkyunkang/lofi-medg.
Current advancements in Multimodal Anomaly Detection (MAD) are largely driven by enhancing multimodal fusion, particularly through the integration of RGB and Depth data for richer anomaly representation. However, less attention was devoted to analyzing the role of cross-modal fusion bias, a well-known challenge in multimodal learning, in MAD. This gap motivates a key question: can we overcome this bias to break the performance bottleneck of current work? In this paper, we first analyze the impact of cross-modal fusion bias in MAD via the Fisher Information Matrix. Then, grounded in these findings, we propose UCFB, a simple yet effective plug-and-play framework designed to mitigate cross-modal fusion bias in MAD. It achieves this by jointly employing Fisher-information-guided dynamic calibration to adjust modality-specific regularization weights and canonical similarity analysis to improve inter-modal interactions. Extensive experiments on the MVTec 3D-AD and Eyecandies datasets demonstrate that UCFB achieves consistent improvements in single-class, multi-class, and few-shot settings.
Zero-shot image captioning aims to generate image descriptions without annotated image-text pairs. Recent approaches exploit text-to-image models to synthesize training data from text-only corpora, but most focus on improving overall data quality. In contrast, we observe that synthetic image-text misalignment is often structured and fine-grained: pairs may remain globally plausible while containing missing entities or misgrounded attributes, thereby degrading supervision fidelity. As a result, methods based on global similarity for image rematching or regeneration may improve apparent plausibility, but cannot systematically repair entity-level misalignment. To address this issue, we propose ReCap, a plug-and-play framework that shifts synthetic data refinement from implicit global matching to explicit fine-grained realignment. Specifically, ReCap enforces entity-level correspondence by using detected image-supported entities to guide caption rewriting, yielding more faithful synthetic supervision. In addition, we introduce an adaptive dynamic weighted learning strategy to downweight unreliable synthetic pairs during training. As a general framework, ReCap can be integrated into existing synthetic-data pipelines. Extensive experiments show that ReCap consistently improves image-text consistency and achieves state-of-the-art performance on both in-domain and cross-domain zero-shot image captioning benchmarks.
In an age of rapid model turnover, how do we make hallucination evaluation more perennial? We explore whether human-written hallucination samples could take the place of model-generated hallucinations, in order to make benchmarking detection independent of particular models. To this end, we construct a dataset of 1,600 human-written samples, spanning four languages (Chinese, English, French, Italian), and 18,400 samples from five vision-and-language models, all annotated for hallucinations using a fine-grained span-level labeling scheme. We find that human-written samples result in higher agreement and allow greater control of dataset contents, while remaining distributionally similar to samples derived from vision-and-language samples and providing a reasonable portrayal of detection capabilities - suggesting that human data is a viable substitute for model-based hallucination benchmarks.
Constructing photorealistic Free-Viewpoint Videos (FVVs) of dynamic scenes from a set of posed 2D images has been an intriguing yet challenging task in computer vision. Methods based on neural rendering achieve high-fidelity image quality in FVV construction. However, most of these methods are unable to achieve real-time rendering and often require complete video sequences to train. Despite the existence of some online training methods capable of rendering FVVs in real time, they struggle to meet the requirements for storage and training time for downstream applications. To overcome this problem, we propose Struct-GStream, which can achieve efficient FVV streaming using structured 3D Gaussians (3DGs). Specifically, we introduce dynamic anchor points to generate structured 3DGs to construct basic scenes and model approximate scene movements based on the assumption of local rigidity in object motion. Besides, we introduce a global free 3DGs patching strategy involving free 3DGs' generation, pruning, and optimization to patch and model deficient areas and emerging objects. Our method achieves fast training at low bitrates while maintaining high rendering quality. Extensive experiments demonstrate that Struct-GStream significantly outperforms existing online training methods for FVV construction in terms of training time, storage, and rendering quality while maintaining competitive rendering speed.
Multimodal Large Language Models (MLLMs) are increasingly expected to solve structured perception tasks that require visual recognition, language-to-object binding, object cardinality preservation, and precisely localized grounding and segmentation outputs. However, existing group-relative reinforcement learning methods provide only response-level supervision, creating a granularity mismatch for structured multi-object prediction: a single advantage is broadcast to all tokens in a response, without distinguishing individual box contributions. To address this mismatch, we propose MCR-GRPO, a marginal contribution assignment framework that derives box-level credit directly from each sampled response. Specifically, Marginal Contribution Reward (MCR) estimates each predicted box's contribution through a leave-one-out comparison, measuring how the matched set value changes when the box is removed from the response. After within-response normalization, records that improve the set value receive positive credit, while redundant or harmful ones are suppressed. To make marginal attribution stable and informative, we further introduce a Continuous Matched Set Value Evaluator that integrates permutation-invariant matching, count-aware normalization, and graded localization. MCR-GRPO maps normalized box-level marginal advantages to the token spans that generated each box, preserving GRPO's response-level comparison while enabling box-aware optimization of structured multi-object grounding. Experiments across REC, DOD, segmentation, and counting benchmarks show state-of-the-art performance over prior GRPO-based baselines.
Extended KAFR: A kinematic-adaptive paradigm for the efficient analysis of surgical video
扩展KAFR：一种用于高效分析外科视频的运动自适应范式
Nguyen, Huu Phong, Khairnar, Shekhar Madhav, Sankaranarayanan, Ganesh
Abstract
Artificial Intelligence is increasingly applied to surgical video analysis for phase segmentation, skill assessment, and workflow optimization. A key challenge is the length of surgical recordings, often one to several hours, creating substantial computational burden. We previously developed Kinematics-Adaptive Frame Recognition (KAFR) for robotic surgery, showing that tracking tool motion effectively identifies informative frames while filtering redundant content. However, laparoscopic surgery introduces additional challenges: manual camera control causes frequent motion artifacts, and image quality is generally lower than robotic systems. This study evaluates whether KAFR generalizes to laparoscopic surgery using the Cholec80 benchmark, comprising 80 laparoscopic cholecystectomy procedures annotated for seven surgical phases. KAFR operates in three stages: a fine-tuned YOLO model detects and segments surgical tools; frames are adaptively selected based on tool displacement or velocity variation; and an X3D model classifies selected frames into surgical phases. KAFR achieved a 91.0\% F1 score using only 0.58\% of frames for phase classification, representing an approximately seven-fold reduction compared to typical 4\% frame sampling, while maintaining performance comparable to LoViT (90.2\%) and Trans-SVNet (89.7\%). These results demonstrate that kinematics-based frame selection transfers effectively to the challenging laparoscopic environment.
One Query, Many Scales: Sparse Mixture-of-Experts for Efficient Hierarchical Cross-View Geo-Localization
一个查询，多种尺度：稀疏专家混合模型用于高效的层次交叉视图地理定位
Fan, Ruijie, Ye, Junyan, Zhu, Qi, Li, Weijia
Abstract
Cross-view geo-localization (CVGL) retrieves geo-tagged satellite imagery for a ground-view query. Most systems exhaustively search a flat, fixed-resolution gallery, incurring high cost over large areas and adapting poorly to satellite resolution changes. Autoregressive coarse-to-fine alternatives reduce comparisons but bind later predictions to earlier decisions and a predefined hierarchy. We introduce GeoMoE, a sparse mixture-of-experts dual encoder that decouples global multi-scale representation learning from local hierarchical search. Global multi-scale supervision and content-adaptive routing map ground and satellite images across resolutions into a globally comparable embedding space. At inference, each image is encoded once, and probabilistic beam search follows parent--child links to score a small candidate subset. Later levels reuse these descriptors rather than features generated by preceding levels, limiting feature-level error propagation and hierarchy coupling. We further introduce VIGOR-M, a four-city benchmark with an explicit parent--child satellite hierarchy and held-out half-step galleries for single-resolution, cross-resolution, and hierarchical evaluation. GeoMoE achieves 95.78% R@40m on Just Zoom In, 2.77 percentage points above the previous best, and 62.39% R@1 on VIGOR-M. The latter requires 0.885 MMAC/query for descriptor matching, 5.27% of an exhaustive L3 scan, while exceeding the strongest exhaustive baseline by 3.12 percentage points in R@1. One model trained on L1, L2, and L3 also outperforms a matched dense control across all six galleries and transfers to three withheld resolutions. By decoupling globally trained embeddings from local hierarchical search, GeoMoE jointly improves localization accuracy, search efficiency, and cross-resolution transfer.
Training-free test-time defenses offer a practical way to improve the adversarial robustness of CLIP-style vision--language models without modifying the pretrained model. However, their correction strength is typically fixed for a narrow range of attack budgets, even though the attack budget is unknown at inference and the required correction varies across samples. We show that this mismatch causes existing defenses to degrade sharply as attacks strengthen. We introduce ReACT-CLIP, a response-conditioned test-time defense that separately determines how strongly each input should be corrected and whether defensive intervention is necessary. Our key observation is that the relative increase in CLIP visual-feature drift between low- and high-noise probes provides a graded, sample-specific proxy for correction demand. ReACT-CLIP maps this relative cross-noise drift to the Gaussian noise scale used to construct a stable, noise-averaged feature anchor, enabling the corrective reach to adapt to each input. To determine whether intervention is necessary, we further observe that clean inputs retain stable class-probability distributions under weak spatial augmentations, whereas adversarial inputs exhibit greater variation. ReACT-CLIP quantifies this variation using a prediction-instability score computed by Jensen--Shannon divergence and combines it with relative cross-noise drift to form the defensive intervention score. ReACT-CLIP requires no model or prompt training, and its correction-strength mapping is calibrated once and fixed across datasets and attack budgets. Across 12 downstream datasets, as well as ImageNet and its distribution-shifted variants, ReACT-CLIP delivers substantial robustness gains across diverse attack types and strengths while largely preserving clean accuracy.
PlantRig - From Bones to Branches: Adaptation of Autoregressive Rigging Models for Plant Skeletal Reconstruction
PlantRig - 从骨骼到分支：自回归绑定模型在植物骨架重建中的适应性
Hu, Nathan, Yang, Yang, Okura, Fumio
Abstract
Autoregressive rigging models such as UniRig and SkinTokens perform well on articulated characters, but their ability to generalize to plant structures remains largely unexplored, since plant topologies exhibit highly variable, non-canonical branching patterns that challenge learned skeletal priors. We evaluate these models for plant skeletal reconstruction using synthetic L-system-generated trees and real scanned data spanning monopodial, sympodial, whorled, and vine-like archetypes. Preliminary testing showed UniRig collapsing complex branching into near-linear chains, while SkinTokens preserved topology better but over-segmented branches and produced an unstable output space, so we focused on UniRig for its greater stability. Diagnosis traced the collapse to sampling-level suppression of branch tokens, and further analysis showed the frozen mesh encoder had limited sensitivity to structural variation, pointing to a geometric bottleneck in the tokenization pipeline rather than a purely learned bias. Building on these findings, we applied multi-round fine-tuning over multiple procedurally generated synthetic datasets. Across rounds, the model progressively recovered accurate branching topology and generalized beyond branch-only structures to plants with foliage, a harder case given the zero-thickness, mesh-normal-dependent geometry of leaves. The resulting model generalized well across diverse plant forms without leaf-specific architectural changes, indicating that targeted fine-tuning can substantially close the domain gap between character-rigging priors and plant skeletal structure. As such, our work points toward a viable path for automated plant rigging across both branch topology and foliage type, even those not considered in our findings.
Lethe: How Hard Is It to Forget? A Benchmark for Federated Unlearning in Medical Imaging
Lethe：忘记有多难？医学影像中的联邦去学习基准
Chen, Shengchao, Shu, Ting
Abstract
Federated learning enables medical-imaging models to be trained across hospitals, and privacy law, most explicitly the GDPR ``right to be forgotten'', turns removing a hospital's, a class's, or a patient's influence from such a model into a federated unlearning problem. This need is most acute in medicine, where patients withdraw consent and hospitals leave collaborations. Yet nearly all unlearning evidence comes from natural images, whose heterogeneity and task structure differ sharply from clinical data, so it is unclear whether existing methods transfer, and no shared protocol covers clinical data. We present Lethe, a benchmark for federated unlearning in medical imaging. It evaluates twelve methods across eight task families, from classification and segmentation to denoising, cross-modality synthesis, and vision-language question answering, at three forgetting granularities and against a retrained gold standard on utility, privacy, and cost. The central result is that what separates methods is the difficulty of the forgetting request, not the method itself. The easy removals that dominate the literature leave the methods that preserve utility indistinguishable, while only hard ones separate them. More striking, on the many medical tasks that generalize across sites, forgetting a client barely changes task performance, leaving residual membership as the signal that must be erased.
SSR: Similarity-Shift Refinement for Training-Free Object-Centric Masks
SSR：无训练的对象中心掩膜相似性偏移精炼
Lu, Xiaoqian, Guo, Guangfu
Abstract
Object-centric models often produce fragmented masks, boundary leakage, and incorrect region merging. We introduce Similarity-Shift Refinement (SSR), a training-free post-hoc method for improving object-centric masks with a frozen self-supervised Vision Transformer. SSR measures changes in pairwise patch similarity before and after self-attention value aggregation, retains positively strengthened relations, and constructs a sparse affinity graph. This graph propagates the initial soft slot assignments in a single refinement step, without retraining or modifying either model. Across natural-image, synthetic-video, and real-world-video benchmarks, SSR improves all-pixel Adjusted Rand Index in all 24 evaluated model-dataset combinations, with an average gain of 8.5 percentage points. Ablations show that value-space similarity shifts outperform query- and key-space variants as well as static Transformer affinities. However, texture-dense scenes may cause visually similar regions to be over-grouped. Overall, SSR provides a simple and transferable signal for training-free object-centric mask refinement.
Whole slide image (WSI) classification is an evidence-driven task, where diagnostic cues are often sparse, spatially organized, and class-dependent. Existing MIL and vision-language methods aggregate a large pool of patch features into a single global slide representation. Under few-shot supervision, limited slide-level labels make it difficult to learn a reliable aggregation mechanism that organizes sparse local cues into compact and coherent diagnostic evidence. Moreover, a shared slide representation compresses evidence supporting a candidate class and its alternatives into the same feature, limiting class-specific reasoning and interpretability. To address these issues, we propose EviBall, a class-conditioned evidence retrieval framework for few-shot WSI classification. EviBall organizes local patches into Evidence Balls through semantic-spatial assignment and center refinement, yielding compact and spatially coherent evidence units under weak supervision. It then uses task-specific class queries, including language-guided queries for morphology-oriented tasks and molecular-guided queries for molecular endpoint prediction, to retrieve supporting evidence balls and produce class-conditioned evidence representations for direct class-wise prediction. By introducing structured evidence units and task-relevant semantic guidance, EviBall reduces the reliance on learning an unconstrained global aggregation mechanism from scarce slide-level labels. It therefore reformulates few-shot WSI classification as structured evidence retrieval and competition among candidate classes. Extensive experiments across four morphology-oriented and molecular endpoint WSI tasks demonstrate that EviBall consistently outperforms conventional and vision-language MIL baselines under diverse few-shot settings, while providing spatially localized and class-specific evidence for each prediction.
Understanding and generating spatially coherent layouts from natural language remains a fundamental yet challenging task for large language models (LLMs). Existing LLMs often struggle to capture explicit geometric relationships and structural dependencies between objects. To address this issue, we propose SG-Layout, a graph-guided layout generation framework that explicitly incorporates structured spatial knowledge into LLMs. SG-Layout follows a two-stage training paradigm: (1) a graph-language feature alignment stage, where a relational graph encoder and a projector are trained to map scene-graph embeddings into the LLM's linguistic space; and (2) an instruction tuning stage, where LoRA-based adapters enable efficient fine-tuning for instruction-driven layout generation while keeping the backbone frozen. We evaluate SG-Layout on image layout generation, indoor scene synthesis and robotic object rearrangement tasks. Experimental results show that SG-Layout improves spatial reasoning accuracy and geometric consistency over the compact open-source backbone, with particularly clear advantages in relation-dense and compositionally complex scenes. These results highlight the effectiveness of graph-structured feature alignment for enhancing controllable layout generation.
Liang, Sen, Guan, Fengbin, Zhang, Youliang, Li, Xin, Chen, Zhibo
Abstract
Text-driven instruction-based video editing in complex scenes remains challenging: purely textual prompts often fail to capture precise spatial relationships and physical constraints, resulting in target ambiguity and physically implausible outcomes. To address this, we propose a plan--guide--edit framework that explicitly bridges semantic intent and spatial execution. In our framework, a Chain-of-Thought (CoT)-enhanced multimodal large language model (MLLM) serves as a planner, performing structured reasoning over the video and instructions to derive a precise sequence of bounding boxes and attribute-enriched editing directives. These spatial priors then guide a box-conditioned mask generator, transforming ambiguous global retrieval into localized, context-aware refinement and producing masks that more accurately capture object scale, contact relationships, and placement. Building on these spatial and semantic signals, a diffusion-based editor integrates the masks, enriched instructions, and frame features to render high-fidelity edits that remain temporally coherent and spatially well aligned. Trained first in a modular manner and then jointly, our framework achieves superior performance with reduced data requirements, delivering precise localization in scenes with multiple similar objects and physically consistent object additions, and extensive experiments demonstrate state-of-the-art performance over multiple strong baseline methods. More details are available at: https://github.com/flying-sky999/CoT-Edit
Chinese Translation
基于文本驱动的复杂场景指令视频编辑仍然面临挑战：纯文本提示往往无法准确捕捉空间关系和物理约束，导致目标模糊和不符合物理规律的结果。为了解决这个问题，我们提出了一种计划-引导-编辑框架，明确连接语义意图与空间执行。在我们的框架中，增强链式思维（Chain-of-Thought, CoT）的多模态大型语言模型（Multimodal Large Language Model, MLLM）作为规划者，对视频和指令进行结构化推理，以推导出精确的边界框序列和丰富属性的编辑指令。这些空间先验信息随后引导一个基于框的掩膜生成器，将模糊的全局检索转化为局部的、上下文感知的细化，并生成更准确捕捉对象尺度、接触关系和放置位置的掩膜。在这些空间和语义信号的基础上，基于扩散的编辑器整合掩膜、丰富的指令和帧特征，以渲染高保真编辑，保持时间一致性和空间良好对齐。我们的框架首先以模块化方式训练，然后联合训练，取得了优越的性能，减少了数据需求，在多个相似对象的场景中实现了精确定位和物理一致的对象添加，广泛实验表明其在多个强基线方法上达到了最先进的性能。更多细节可见：https://github.com/flying-sky999/CoT-Edit
Video world models predict future observations conditioned on historical observations and control signals, enabling long-horizon generation through autoregressive state transitions. Unlike conventional video generation models that primarily capture visual appearance and motion, video world models learn the underlying dynamics governing environment evolution under agent actions, providing a foundation for embodied AI and interactive simulation. Recent progress has largely relied on adapting pretrained video generation models through post-training or distillation. Although effective, these approaches often require complex training pipelines, substantial computational resources, and suffer from the mismatch between bidirectional pretraining and causal streaming inference. Recent studies have shown that training autoregressive video world models from scratch is feasible and scalable. However, the community still lacks a lightweight, transparent, and fully reproducible baseline trainable end-to-end with modest computational resources. We present MiniWorld, a reproducible framework for training streaming video world models from scratch. MiniWorld employs a block-causal Video Diffusion Transformer trained with Flow Matching in the latent space of a pretrained Video VAE. Building on Diffusion Forcing, it adopts a chunk-wise non-decreasing noise schedule and two-stage continued training to improve temporal modeling and stability. During inference, MiniWorld combines a rolling KV cache with pipelined asynchronous denoising for efficient streaming generation under bounded computation. The entire model can be trained within several days on a single 8-GPU server. By releasing the training and inference codebase and pretrained checkpoints, we hope MiniWorld will facilitate future research on video world modeling.
Large language models have made text the default medium for human--AI interaction, buttext alone cannot express the full range of responses required by multimodal assistants,avatars, and embodied agents. While recent audio-video generative models can synthesizehigh-fidelity synchronized content, existing supervision is largely \emph{descriptive}:models are trained to render captions rather than to produce audio-visual responsescaused by external user interactions. We introduce \textbf{InteracVid}, \emph{the firstopen-source large-scale dataset that addresses this missing supervision}, so that everysample couples a preceding audio-visual context and an external stimulus with the realinteractive response that follows. We design a metadata-aware pipeline that extractsinteractive clips from long, noisy livestreams, yielding over \textbf{454K}context-query-response triplets from more than \textbf{59K} livestream videos andspanning conversation-centered, object-centric, procedural, embodied, and screen-basedscenarios. A ten-rater human study confirms that the extracted interactions are causal,natural, and temporally complete for both genuine and reconstructed queries. On aheld-out benchmark of \textbf{100} genuine live-chat queries, fine-tuning on InteracVidimproves both interaction planning and audio-video response generation, and anindependent human evaluation reproduces the system ranking and the conclusions obtainedwith our automatic judge. These results highlight interaction-structured data as acritical foundation for interactive multimodal generation.
Think in Sets for Streaming Video Token Compression
基于集合思维的流媒体视频令牌压缩
Duan, Moxu, Fu, Jingwen, Wang, Yuwang
Abstract
Streaming VideoLLMs process frames causally while visual tokens grow continuously, making compression essential for controlling prefilling latency and memory. Existing training-free methods independently rank tokens, ignoring marginal-gain interactions among retained tokens. We argue that streaming video token compression should instead be formulated as set selection, where each candidate is valued by what it adds beyond the tokens already retained. Unlike existing set-wise methods designed for offline tasks, streaming makes causal, frame-by-frame pruning decisions, so modeling cross-frame interactions requires an explicit historical reference. This creates a reference-set dilemma: the reference must adequately represent previously conveyed content while remaining bounded for real-time inference. We introduce NovaCov, to our knowledge the first training-free, plug-and-play set-wise token compressor designed for streaming video. NovaCov maintains a capacity-bounded, recency-weighted Historical Reference Bank and optimizes a dual-branch submodular coverage objective that preserves representative current-frame content while prioritizing information insufficiently covered by history. Both branches are facility-location functions, so greedy selection retains the classical (1-1/e) approximation guarantee. Across streaming and offline benchmarks, NovaCov outperforms existing training-free compression methods, retaining 99.6% of ReKV accuracy while reducing LLM prefilling latency by 46%.
We present DynActiveGS, a dynamic-aware active reconstruction framework based on 3D Gaussian Splatting (3DGS) for autonomous exploration in dynamic environments. The framework incrementally reconstructs a 3D Gaussian scene representation while suppressing motion-corrupted observations through online uncertainty prediction and uncertainty-weighted Gaussian optimization. A key component of DynActiveGS is the explicit decomposition of uncertainty into structural uncertainty and motion-induced uncertainty, which enables the system to distinguish under-reconstructed static regions from dynamically unreliable areas. Based on these uncertainty fields, DynActiveGS performs dynamic-aware viewpoint selection and dynamic-constrained path planning to favor informative yet stable observations during exploration. The resulting system forms a unified closed-loop pipeline for robust active reconstruction in dynamic scenes. Extensive experiments on challenging dynamic benchmarks demonstrate consistent improvements over existing active reconstruction baselines in reconstruction accuracy, completeness, rendering quality, and exploration efficiency.
3DZip: Spatial-Aware Feature Diversity-Guided Token Compression for 3D Question Answering
3DZip：基于空间感知特征多样性的引导令牌压缩用于3D问答
Baek, Changwoo, Kong, Kyeongbo
Abstract
Recent 3D vision-language models (3D VLMs) construct geometry aware tokens by projecting 2D visual features into world coordinates, enabling spatial reasoning for tasks such as 3D question answering. However, this design generates thousands of tokens per scene, resulting in substantial computational and memory overhead. While token compression has been extensively studied in 2D VLMs, existing approaches rely on semantic relevance or attention-based selection that overlook the structured spatial nature of 3D tokens. Moreover, redundancy in 3D representations cannot be resolved by spatial proximity alone, as object-level token imbalance persists even after spatial aggregation. To address this, we propose 3DZip, a three-stage token compression framework that first applies coarse voxelization to remove point-level redundancy, then selects anchor tokens based on feature-space diversity via a Determinantal Point Process, and finally merges remaining tokens under spatial constraints to preserve geometric coherence. Experiments on three 3D question answering benchmarks demonstrate that 3DZip consistently outperforms existing compression methods, retaining 94.7% of the original performance with only 128 tokens, achieving a $1.92\times$ faster inference speed.
Chinese Translation
近期的3D视觉语言模型（3D VLMs）通过将2D视觉特征投影到世界坐标中构建几何感知令牌，从而实现对3D问答等任务的空间推理。然而，这种设计为每个场景生成了数千个令牌，导致了显著的计算和内存开销。尽管在2D VLMs中对令牌压缩进行了广泛研究，但现有方法依赖于语义相关性或基于注意力的选择，忽视了3D令牌的结构化空间特性。此外，仅通过空间邻近性无法解决3D表示中的冗余问题，因为即使在空间聚合后，物体级令牌的不平衡仍然存在。为了解决这一问题，我们提出了3DZip，一个三阶段的令牌压缩框架，首先应用粗略体素化以去除点级冗余，然后通过行列式点过程（Determinantal Point Process）基于特征空间多样性选择锚令牌，最后在空间约束下合并剩余令牌以保持几何一致性。在三个3D问答基准上的实验表明，3DZip始终优于现有的压缩方法，仅用128个令牌保留了94.7%的原始性能，实现了$1.92 imes$的推理速度提升。
While feed-forward 3D Gaussian Splatting (3DGS) enables efficient 3D reconstruction, achieving high-fidelity rendering remains challenging. Existing pixel-aligned approaches suffer from spatial inflexibility and massive structural redundancy, whereas query-based methods lack 3D priors and entangle geometry with appearance, yielding blurry, pose-dependent results. To overcome these deficiencies, we propose \textbf{QuerySplat}, a feed-forward 3DGS framework driven by geometric priors and explicit appearance decoupling. Specifically, we design a dual-branch query-based decoder: the geometry branch leverages a pretrained Vision Geometric Model for spatial understanding, which intrinsically endows QuerySplat with pose-free modeling capabilities, while the appearance branch recovers high-frequency details through a dedicated pathway separated from geometric attribute regression. Extensive experiments demonstrate that QuerySplat mitigates the blurry rendering issues of early query-based models and consistently outperforms pixel-aligned approaches in rendering fidelity. On the challenging DL3DV benchmark, it achieves state-of-the-art novel view synthesis performance, with average PSNR gains of 2.30 dB and 1.04 dB over the best pose-free and pose-required baselines, respectively. Project Page: https://inspatio.github.io/querysplat.
Fruit-HSNet: A Machine Learning Approach for Hyperspectral Image-Based Fruit Ripeness Prediction
Fruit-HSNet：一种基于高光谱图像的水果成熟度预测的机器学习方法
Jmaa, Ahmed Baha Ben, Chaieb, Faten, Fabijańska, Anna
Abstract
Fruit ripeness prediction (FRP) is a classification-based agricultural computer vision task that has attracted much attention, thanks to its wide-ranging advantages in agriculture field for both pre-harvest and post-harvest management. Accurate and timely FRP can be achieved using machine/deep learning-based hyperspectral image classification techniques. However, challenges including the limited availability of labeled data and the lack of robust methods generalizable to various hyperspectral cameras and fruit types can compromise the effectiveness of hyperspectral image-based FRP. Addressing these challenges, this paper introduces Fruit-HSNet, a machine learning architecture specifically designed for hyperspectral classification of fruit ripeness. Fruit-HSNet incorporates a spatio-spectral feature extraction module based on Fourier Transform and central pixel spectral signature followed by learnable feature fusion and a classifier optimized for ripeness classification. The proposed architecture was evaluated using the DeepHS Fruit dataset, the largest publicly available labeled real-world hyperspectral dataset for predicting fruit ripeness, which includes five different types of fruits-avocado, kiwi, mango, kaki, and papaya-captured with three distinct hyperspectral cameras at various stages of ripeness. Experimental results highlight that Fruit-HSNet substantially outperforms existing deep learning methods, from baseline to state-of-the-art models, with improvements of 12%, achieving a new state-of-the-art overall accuracy of 70.73%.
It's the Decoding Format, Not the Perturbation: Auditing Consistency-Based Selection for Vision-Language Test-Time Scaling
解码格式而非扰动：审计基于一致性的视觉-语言测试时缩放选择
Zheng, Puzhuo, Kurban, Hasan
Abstract
Test-time scaling lifts large language model reasoning by sampling many candidate solutions and selecting among them, yet the same recipe transfers poorly to vision-language models (VLMs): recent work shows that simple majority voting beats selection methods built on the model's own self-verification, apparently because at the selection layer an image-grounded answer and a confident guess from the language prior look the same. A natural fix is to make the selection signal one that cannot be computed without the image. We study Perturbation Grounded Selection (Pgs), a label-free, training-free rule that scores each candidate by whether the model re-derives it under label-preserving perturbations of the input (cropping, background masking, mild photometric or geometric jitter); Pgs recovers majority voting when the perturbation set is empty. The decisive question is not whether Pgs beats chain-of-thought only majority voting, but whether the perturbation term adds anything once decoding format and budget are controlled. We therefore introduce a format-matched control (MatchedCtrl): the same short, no-CoT draws spent on the original image. Across TextVQA, MATH-Vision, MMMU, and ViLP, with a Qwen headline (three-seed means) and LLaVA-OneVision coverage in matched-budget selector tables, Pgs appears to beat plain majority voting by up to +31.8 points on TextVQA (Qwen), but MatchedCtrl tracks or exceeds Pgs within noise on every benchmark, including the vision-required ViLP; no Qwen category shows a significant gain over this control. The stability gap is real and image-dependent (up to +0.48), yet does not predict per-instance wins. The result is negative and diagnostic: perturbation consistency is at best a partial diagnostic of visual dependence and, on its own, not a usable selection signal once format is controlled; gains reported against CoT-only majority voting overstate such methods.
Visual document retrieval has recently become increasingly important in applications such as enterprise search, scientific literature discovery, and retrieval-augmented generation. These applications depend on efficiently identifying query-relevant pages across large collections of visually rich documents. Existing methods commonly adopt late-interaction architectures that encode and index documents offline to enable scalable and low-latency online retrieval. Despite its efficiency, this paradigm requires each document to be encoded into a fixed representation before the query is known. However, the same content in a visual document may induce different interpretations depending on the query intent, which a fixed representation struggles to capture. Yet postponing document encoding until the query arrives would incur prohibitive online retrieval latency. To address this gap, we propose VaRS-Doc, a visual document retrieval framework that diversifies document representations by enabling the model to actively explore variant latent interpretations during document encoding, while preserving efficient late-interaction retrieval in which each query adaptively selects the best-fit representation. We further introduce a two-stage training strategy that encourages the model to capture complementary semantic interpretations and prevents it from falling back to train a single dominant representation. Experiments on visual document retrieval benchmarks show that VaRS-Doc achieves state-of-the-art retrieval performance, offering a practical solution to the mismatch between query-agnostic document encoding and query-specific retrieval needs. Code is available at https://github.com/bokufa/VaRS-Doc.
From Forest to Future Capital: Tracking Land Cover Change in Ibu Kota Nusantara (IKN) from 2021 to 2026 with PlanetScope Imagery
从森林到未来资本：利用PlanetScope影像追踪Ibu Kota Nusantara (IKN) 2021至2026年间的土地覆盖变化
Ong, Clarissa Rui Min, Loo, Elizabeth Tee Inn, Soh, Kenneth Woon Hao, Rachmadi, William, Zheng, Qiming, Li, Hao
Abstract
Indonesia's relocation of its political and administrative capital from Jakarta to Ibu Kota Nusantara (IKN) has been framed around a ``Forest City'' vision, yet rapid construction within the Core Government Area (KIPP) raises concerns over land conversion, vegetation loss, and carbon stock decline. This study applies remote sensing techniques to systematically assess land use and vegetation cover change in KIPP from 2021 to 2026 using PlanetScope SuperDove satellite imagery. Cloud-free mosaics were prepared and analysed through spectral indices, including the Normalised Difference Vegetation Index (NDVI), Normalised Difference Red Edge (NDRE), and Normalised Difference Water Index (NDWI), alongside supervised land use and land cover (LULC) classification using a Support Vector Machine algorithm. Results show substantial environmental transformation, with mean NDVI declined by 17.1\%, total carbon stock decreased by 0.28\%, developed land expanded by 672\%, and total vegetation declined by 18.1\%. Vegetation loss was most extensive between 2023 and 2024, although a temporary recovery in NDVI and carbon stock occurred from 2024 to 2025 as active clearing slowed and development shifted towards already-cleared land. Overall, the findings demonstrate that remote sensing provides an effective approach for monitoring the environmental impacts of large-scale urban development, while highlighting the need for higher-resolution, hyperspectral, and SAR-based methods to improve detection of construction stages and plantation-related land cover changes.
Chinese Translation
印度尼西亚将其政治和行政首都从雅加达迁至Ibu Kota Nusantara (IKN)的计划围绕着“森林城市”的愿景展开，但核心政府区（KIPP）内的快速建设引发了对土地转化、植被损失和碳储量下降的担忧。本研究应用遥感技术，系统评估2021至2026年间KIPP的土地利用和植被覆盖变化，使用PlanetScope SuperDove卫星影像进行分析。研究准备了无云的马赛克影像，并通过光谱指数进行分析，包括归一化差异植被指数（NDVI）、归一化差异红边指数（NDRE）和归一化差异水体指数（NDWI），同时使用支持向量机算法进行监督土地利用和土地覆盖（LULC）分类。结果显示环境发生了显著变化，平均NDVI下降了17.1%，总碳储量减少了0.28%，开发土地扩大了672%，总植被减少了18.1%。植被损失在2023至2024年间最为严重，尽管在2024至2025年间，由于清理活动减缓且开发转向已清理土地，NDVI和碳储量出现了暂时恢复。总体而言，研究结果表明遥感为监测大规模城市开发的环境影响提供了一种有效的方法，同时强调了提高分辨率、使用高光谱和合成孔径雷达（SAR）方法的必要性，以改善对建设阶段和种植相关土地覆盖变化的检测。
The realism of images generated by multimodal large language models (MLLMs), such as GPT Image2 and Nano Banana2, has improved rapidly in recent years. Compared with early generative models, current models have made clear progress in text rendering. They can produce high-quality images that closely resemble real-world application scenarios. The enhanced generation capabilities of current MLLMs pose increasingly severe challenges to AI-generated image detection. Detection is no longer limited to identifying obvious artifacts left by early generators. Instead, it requires systematic and realistic benchmarks for the new generation of generated content. However, most existing benchmarks are still built around early generative models and cannot fully evaluate the forensic challenges introduced by high-quality and multi-form generated images. To address this gap, this paper constructs a benchmark dataset for detecting images generated by MLLMs. The benchmark covers several realistic application scenarios and adopts three generation protocols to simulate direct generation, reference-based reconstruction, and local editing. Based on this benchmark, we evaluate detector degradation from traditional scenarios to MLLM-generated images and analyze false positive rates and false negative rates across three sample types, revealing the failure modes of existing methods. We further propose a structural-artifact-prior-guided dual-stream prompt framework (SAP-DSP) as a strong baseline. SAP-DSP uses dual-stream prompt learning and structure-aware routing fusion to improve representation learning. Extensive experiments show that the proposed benchmark exposes the performance degradation of existing detectors on high-quality generated images, while SAP-DSP achieves more stable detection results on this benchmark. Our code and dataset are publicly available at https://github.com/xbrainnet/SAP-DSP.
Video large language models (Video-LLMs) represent videos as dense sequences of visual tokens, whose length grows with the temporal and spatial extent of the input. These tokens often contain substantial redundancy arising from repeated visual patterns, leading to unnecessary computation in the subsequent language-model processing. Existing token compression methods, including pruning and merging, perform compression online during inference, repeatedly incurring additional computation for each input video and often relying on model-specific designs that limit their generality, we instead rethink this paradigm by shifting the costly compression process offline. We propose \textbf{ONCE}, a plug-in video token compression framework that introduces an offline-to-online paradigm: a frequency-aware global codebook is learned once in the visual feature space and reused for lightweight online compression through codebook lookup and aggregation, reducing repeated per-video computation and the need for model-specific compression designs. Extensive experiments across multiple video understanding benchmarks and against diverse compression baselines demonstrate that our approach achieves a strong accuracy-efficiency trade-off, maintaining competitive performance while achieving the lowest inference latency among compared methods.
Full-body capture from unconstrained photographs requires global correspondence across arbitrary views, poses, crops, and occlusions. Yet pose, geometry, and foundation features estimated in this setting are too unreliable for dense matching or appearance transfer, while diffusion rectifiers and optimization pipelines expose no common interface for consuming such uncertain correspondence. Our insight is that correspondence need not be locally accurate: its coarse viewpoint and body layout can still organize how a diffusion prior adapts and guides reconstruction. We introduce \emph{Astrolabe}, a host-portable adapter built on frozen viewpoint-guided spherical maps (SPH). A fixed bounded transform converts SPH into a spatial noise shift, which is matched during prior adaptation and reused during downstream denoising or score-distillation guidance in both pipeline categories. When a rectifier exposes a reference router, the same target/reference SPH additionally supplies coarse compatibility scores to select native appearance features; router-free optimization uses only the shared shift path. Astrolabe therefore follows one SPH--shift--adapt--guide process without dense warping or a learned control branch. Across Puzzle-IOI and 4D-Dress, it improves all reported image metrics in both hosts and all paired Puzzle-IOI geometry metrics; image gains extend to rear views, while 4D-Dress geometry remains stable overall.
Recently, diffusion-based removal methods have achieved promising visual quality in removing both target objects and their associated effects. However, they typically rely on multi-step denoising, leading to high inference cost. Directly applying existing one-step distillation methods is also suboptimal, since their global objectives lack explicit region-wise calibration and may weaken the asymmetric edit-and-preserve behavior required by object-effect removal. To address these challenges, we propose TurboClear, a one-step SDXL-based object-effect removal model. During training, we design Region-Calibrated Distribution Matching (RDM) for region-aware distillation to preserve the teacher model's asymmetric edit-and-preserve behavior. Furthermore, we propose Learnable Spatial Fusion (LSF) for lightweight inference-time fusion. Extensive experiments show that TurboClear significantly improves inference efficiency while maintaining competitive visual quality. TurboClear reduces the computational overhead by up to $40.04\times$ compared to ObjectClear, and by up to $665\times$ against the Flux-based method OmniPaint, all while maintaining comparable or better visual removal quality. Code is available at https://github.com/GuoCalix/TurboClear.
Chinese Translation
近年来，基于扩散的去除方法在去除目标对象及其相关效果方面取得了令人满意的视觉质量。然而，它们通常依赖于多步去噪，导致高推理成本。直接应用现有的一步蒸馏方法也并非最佳选择，因为它们的全局目标缺乏明确的区域校准，可能削弱对象效果去除所需的非对称编辑与保留行为。为了解决这些挑战，我们提出了TurboClear，一种基于SDXL的一步法对象效果去除模型。在训练过程中，我们设计了区域校准分布匹配（Region-Calibrated Distribution Matching, RDM）以实现区域感知的蒸馏，从而保留教师模型的非对称编辑与保留行为。此外，我们提出了可学习空间融合（Learnable Spatial Fusion, LSF）以实现轻量级推理时的融合。大量实验表明，TurboClear显著提高了推理效率，同时保持了竞争力的视觉质量。与ObjectClear相比，TurboClear将计算开销降低了最多$40.04 imes$，与基于Flux的方法OmniPaint相比，降低了最多$665 imes$，同时保持了可比或更好的视觉去除质量。代码可在https://github.com/GuoCalix/TurboClear获取。
UDT: Reconciling U-Nets and Diffusion Transformers with Data-Adaptive Token Reduction
UDT：通过数据自适应令牌减少调和U-Net和扩散变换器
Yun, Junno, Alçalar, Yaşar Utku, Akçakaya, Mehmet
Abstract
Diffusion Transformers (DiTs) have emerged as a core architecture in generative modeling due to their scalability and adaptability to multimodal tasks. DiTs comprise isotropic transformer blocks, and learn representations progressively across depth, where the denoising objective drives later layers to focus on fine-detail reconstruction. This results in degraded representation quality and an imbalanced encoder-decoder behavior. Prior approaches such as representation alignment (REPA) mitigate this by encouraging stronger early representations via training regularization. Alternatively, U-Net-style DiT architectures introduce explicit multi-scale encoder-decoder structures for improved convergence. But they build on standard U-Net wisdom via learnable operators for spatial downsampling, which are not well-suited to transformer architectures, introducing inefficiencies and compatibility issues with components such as cross-attention and representation regularization. In this work, we propose UDT, a U-Net diffusion transformer that combines the representation power of DiTs with the encoding-decoding benefits of U-Nets, through data-adaptive token merging for downsampling and upsampling, while preserving the DiT token dimension. Our baseline UDT architecture outperforms existing U-Net DiTs and achieves performance comparable to REPA across all model sizes. Furthermore, using architectural optimization and REPA, UDT outperforms SiT's 7.9 FID at 1400 epochs (w/o CFG) within 40 epochs (~ 40x faster convergence) for XL model size on 256x256 ImageNet. Finally, it achieves strong image generation performance with CFG, reaching FID of 1.38 (320 epochs) with SD-VAE and 1.35 (500 epochs) with VA-VAE, providing a new backbone for DiTs with strong empirical benefits.
Ranking Image Fusion the Way Humans Do: A Learned Pairwise Preference Metric for Infrared-Visible Fusion Assessment
像人类一样对图像融合进行排名：一种用于红外-可见融合评估的学习成对偏好度量
Liu, Haoran, Liu, Mingzhe, Li, Peng, Zan, Guibin
Abstract
Infrared-visible image fusion (IVIF) has no ideal fused reference, so fusion algorithms are routinely ranked by scalar objective metrics that formalize different proxies for information transfer, structure, or source similarity. These proxies often disagree with the judgment that ultimately matters: given the same sources, which of two fused results does a human prefer? Direct pairwise comparison is an established reference protocol for relative subjective assessment, but its cost grows quadratically with the number of algorithms, which prevents routine use. We present the Learned Perceptual Image Fusion Measure (LPIFM), a source-conditioned model that operationalizes the human A/B/Tie comparison protocol as a repeatable, scalable surrogate. LPIFM jointly observes the infrared source, the visible source, and two fused candidates, and predicts whether candidate A is better, candidate B is better, or the two are perceptually equivalent. Supervision comes from a new dense preference corpus that covers every unordered comparison among a broad pool of fusion methods on the scenes of a public benchmark, labeled under a blinded, randomized, two-stage protocol with expert adjudication. Across scene- and method-generalization settings, LPIFM tracks human pairwise decisions closely and reproduces the tie-aware Bradley-Terry rankings derived from human labels; on full method pools it surpasses the strongest conventional metric by a wide margin in both pairwise accuracy and ranking correlation. We release the annotated preference dataset, together with the LPIFM model weights, source code, and evaluation code, to support preference-aligned IVIF assessment. LPIFM offers a practical instrument for human-aligned method comparison and ranking at scale.
State-of-the-art RGB-Event detectors improve the detection of small, fast-moving objects by combining complementary features from RGB and Event data, yet they typically fuse the two modalities into a unified representation for both localization and classification. Such a task-symmetric design is inconsistent with the intuition that the two modalities should play different roles according to their task-specific strengths. To examine this issue, we conduct a modality-specific evaluation and find that the relative advantage of the two modalities reverses across tasks: Event data are substantially more effective for class-agnostic localization, whereas RGB data provide stronger category evidence within localized target regions. Motivated by this task-dependent asymmetry, we propose an Asymmetric Event-RGB Object Detection Transformer (AERODet). During class-agnostic localization, Scale-wise Uncertainty-aware Reliability Estimation (SURE) calculates the relative reliability of the two modalities from their objectness response heatmaps and accordingly calibrates their contributions when the decoder aggregates multimodal features. Once the candidate boxes are obtained, Task-Decoupled Semantic Refinement (TDSR) decouples classification from localization and uses RGB RoI features for fine-grained classification. Extensive experiments on FRED and NeRDD demonstrate that AERODet achieves state-of-the-art performance. In particular, it surpasses the strongest RGB-Event baseline by 10.7 mAP points on the FRED challenging split.
Latents from vision foundation models (VFMs) are semantically rich and well suited for visual understanding. Recent representation autoencoder methods such as RAE have shown that they can provide promising latent spaces for image generation. However, VFM latents remain difficult to model directly: DiT-generated latents exhibit spectral mismatch with encoder latents, especially in high-frequency components. Our channel-wise spectral analysis further reveals that these high-frequency components are diffusely distributed across latent channels and entangled with semantic information, making the latent space difficult for DiT to model. To address these challenges, we propose SPAE, latent adaptation framework for generation. Specifically, SPAE employs a compact bottleneck to distill stable semantic information while suppressing high-frequency components, thereby improving the alignment between DiT-generated latents and encoder latents. In addition, we apply a channel-wise masking strategy to promote the decoupling of semantic information and high-frequency details across bottleneck channels. Experiments show that SPAE achieves a favorable balance among visual understanding, generation quality, and reconstruction fidelity.
Multimodal large language models (MLLMs) increasingly rely on long chain-of-thought reasoning for complex tasks. However, as reasoning sequences lengthen, models may gradually rely less on visual evidence and more on accumulated textual context, leading to visual forgetting. Existing approaches do not directly constrain how visual evidence is used and maintained along the original reasoning trajectory, leaving long-context visual forgetting insufficiently addressed. To address this issue, we propose Remember-R1, a reinforcement learning framework that mitigates long-context visual forgetting by applying process-level supervision directly on the original reasoning trajectory. Specifically, Remember-R1 introduces rewards that encourage broader coverage of matched visual keywords, stronger persistence of visual dependence in later reasoning steps, and greater focus on question-relevant image regions. Experiments across multiple model scales and diverse multimodal benchmarks demonstrate that Remember-R1 consistently improves reasoning performance. Additional analyses further show that it slows the decline of visual attention during generation, supporting its effectiveness in mitigating long-context visual forgetting.
AI-generated video (AIGV) detection aims to distinguish real videos from AI-generated ones. In practice, detectors trained on existing data often fail to generalize to newly emerging generative models, making this task challenging. Therefore, continual learning (CL) is essential for improving the adaptability. However, CL frameworks for this task remain underexplored. To this end, we propose SphereVideo, a novel CL framework for AIGV detection built on two key observations. First, real videos exhibit a compact feature distribution. Based on this, we encourage real video features to cluster around a real prototype on a hypersphere while repelling AI-generated samples, thereby establishing a decision boundary. This prototype serves as a stable anchor for CL, regulating boundary evolution and mitigating catastrophic forgetting. Second, existing methods tend to rely solely on spatial artifacts as shortcuts. To enhance temporal modeling, we introduce a strategy that models the temporal dynamics of real data at both frame and clip levels. By strengthening real data modeling, this strategy further facilitates learning a real prototype and forming a stable decision boundary. Moreover, we construct a comprehensive and challenging benchmark. Extensive experiments demonstrate that SphereVideo achieves an improved plasticity-stability trade-off, outperforming prior methods by 3.08% on seen data and 4.00% on unseen AI-generated data.
Modern autonomous-driving fleets record far more video than human reviewers can inspect. This motivates the need for an automatic clip triage mechanism, to surface rare and review-worthy clips, so that driving models can be fine-tuned to better handle unideal circumstances. We test a label-free approach that scores clips by the prediction-error "novelty" of a self-supervised joint-embedding predictive architecture (JEPA); a frozen V-JEPA video encoder is paired with a lightweight predictor head to reconstruct masked clip embeddings, and clips whose embeddings are hard to predict are flagged as interesting. Evaluated under a realistic protocol that trains on one dataset and tests against footage from others, this approach appears highly effective. We show that this apparent success is actually a domain-shift consequence: on a fair benchmark drawn from a single dataset, this mechanism collapses to chance and is on par with simple no-training baselines. A lightly supervised probe on the same frozen embeddings results in almost double the average precision, indicating that the bottleneck is indeed the self-supervised objective, rather than the representation. We present this as a study for evaluating the effectiveness of self-supervised learning, where cross-dataset protocols can silently reward domain separation over novelty.
High-definition (HD) maps are essential for autonomous driving systems. In constructing such maps, onboard multi-view camera images, standard-definition maps and satellite images provide crucial information. However, due to the modality and perspective differences among these data sources, existing methods often struggle to effectively align and fuse them, making online HD map construction still challenging. To address these issues, we propose Driver2Map, an online HD map construction model inspired by human drivers. Unlike existing HD map construction models that utilize only two modalities, our Driver2Map can simultaneously exploit three modalities. Specifically, we propose a "two-stage alignment" strategy to reduce spatial misalignment across different modalities. Additionally, we introduce "Pose-Guided BEV Fusion", a BEV (bird's-eye-view) generation module that leverages camera pose information to adaptively weight multi-view features, thereby effectively suppressing cross-view feature overlap during BEV generation. Also, we design a "Pretrained Prior for Map Refinement" module to refine the initial prediction by learning map structure priors, thus improving the HD map prediction under dynamic occlusions. Extensive experiments demonstrate that Driver2Map outperforms existing methods on both IoU and AP metrics.
DeVIT: Low-Power Vision Transformer Acceleration Using Delta Computation
DeVIT：基于增量计算的低功耗视觉变换器加速
Hosseinzadeh, Reyhaneh, Moghaddam, Parham Zilouchian, Modarressi, Mehdi
Abstract
The emergence of transformer-based deep learning models has brought unprecedented performance across various domains, particularly in natural language processing and computer vision. However, deploying these models, especially on resource-constrained devices, poses significant challenges due to their high computational complexity and large memory size and bandwidth requirements. This complexity has led researchers to use low-bit model weights to reduce memory usage and improve efficiency. In addition to reducing processing and memory demands, quantization introduces another useful property: value locality, where the extremely large number of parameters are restricted to a limited range of values. To fully take advantage of this locality, this paper presents DeVIT, an acceleration method for vision transformers that leverages differential computation to enable multiplier-less matrix multiplication.
Prompt-Driven Simulation with Feature Perturbation for Cross-Domain Few-Shot Object Detection
基于提示驱动的特征扰动模拟用于跨域少样本目标检测
Zhuo, Linhai, Cai, Junxi, Qian, Tianwen, Zheng, Qingping, Liu, Yang
Abstract
Data augmentation, which simulates diverse visual variations to expand the source distribution and induce synthetic domain shifts, is a simple yet effective strategy for mitigating severe domain shifts and limited labeled target data in cross-domain few-shot object detection (CD-FSOD). Existing approaches rely on conventional data augmentation, such as Color-Jitter, Mosaic, and background-centric adaptation (e.g., Domain-RAG), which are limited in modeling complex domain shifts and often lead to suboptimal performance. In this paper, we propose PSP-FSOD, a principled framework that integrates prompt-driven domain simulation with feature perturbation regularization to improve generalization in CD-FSOD. To enable controllable domain synthesis, we design a prompt-driven strategy that leverages the visual grounding capability of large VLMs to jointly model foreground and background variations, generating semantically consistent yet domain-diverse training samples. Moreover, we adopt a grounding-aware generation scheme that guides object placement and alleviates semantic-spatial misalignment, thereby improving foreground adaptation. To ensure training stability and robustness, we further introduce a noise-induced feature perturbation mechanism that injects Gaussian noise into multi-scale intermediate features with distribution correction, encouraging consistent predictions under perturbations and reducing reliance on domain-specific cues. Extensive experiments demonstrate that PSP-FSOD produces high-quality domain-diverse supervision and learns domain-invariant representations, consistently improving performance across CD-FSOD benchmarks.
Recent studies develop pixel-level multimodal large language models (MLLMs) that support both Region Segmentation and Region Understanding, extending multimodal interaction from whole images to specific objects and regions. However, these methods face two fundamental challenges. First, the scarcity of high-quality mask--text pairs leaves abundant mask annotations without corresponding language supervision. Second, discrepancies in supervision formats and learning-signal densities induce optimization interference between Region Segmentation and Region Understanding. To address these challenges, we propose PixVL, a self-supervised post-training framework that introduces a unified Mask--Text Consistency Cycle, enabling pixel-level MLLMs to generate and self-verify regional descriptions and learn from unlabeled data. We found that direct cycle based solely on geometric reconstruction is unreliable because re-segmentation IoU does not faithfully reflect the semantic quality and referring sufficiency. PixVL therefore introduces confuser-aware semantic verification, which uses the model's confidence when it correctly chooses the target among highly similar candidate regions, and assigns zero reward to an incorrect choice. Meanwhile, PixVL performs cross-view verification using temporally separated video frames or geometrically transformed image views, preventing cyclic learning from collapsing to positional and shape shortcuts. Finally, a quality-coupled bidirectional learning strategy uses the highest-reward description to guide Text-to-Mask learning. This strategy transforms Region Understanding and Region Segmentation from competing tasks into mutual generators and verifiers. Experiments demonstrate that PixVL improves both region understanding task and segmentation task.
CORTIVA: Candidate-Score Fusion of Complementary Visual Teachers for EEG- and MEG-to-Image Retrieval
CORTIVA：用于脑电图（EEG）和脑磁图（MEG）到图像检索的互补视觉教师候选分数融合
Wang, Junhan, Chen, Kani
Abstract
Decoding visual experience from non-invasive brain activity is central to neuroscience and brain-computer interfaces. Functional magnetic resonance imaging (fMRI) offers fine spatial detail, but its slow hemodynamics and burdensome acquisition limit temporally resolved decoding. Electroencephalography (EEG) and magnetoencephalography (MEG) provide millisecond resolution, making image retrieval compelling: identify the viewed image from one neural response and a fixed candidate bank. Contrastive alignment to pretrained visual representations enables zero-shot retrieval from EEG and MEG, but most systems collapse heterogeneous visual supervision into a single embedding before ranking. This early consolidation imposes one similarity geometry on every candidate order and removes encoder-specific disagreements from the final ranking. We propose CORTIVA, a candidate-score fusion framework that preserves this complementary evidence. Three decoding routes are aligned to heterogeneous visual targets, score the same indexed candidates independently, and combine only their temperature-scaled score vectors before ranking. On the 200-way THINGS-EEG2 benchmark, CORTIVA reaches 73.5% Top-1 and 95.3% Top-5 across ten participants, exceeding the strongest reported baseline by 10.3 and 5.4 percentage points. With a modality-specific neural encoder, the same fusion principle reaches 42.4% Top-1 on THINGS-MEG. Matched route-removal retraining and four weight controls demonstrate that CORTIVA's gain arises from integrating complementary route scores and persists with uniform weighting, without requiring a specialized weighting rule. Independent DINOv2 analyses further reproduce the local error neighborhoods and posterior neural-visual correspondence. These results establish candidate-score fusion as a simple and testable alternative to embedding-level consolidation for neural image retrieval.
Pathology foundation models (PFMs) provide strong tissue representations and have become central to digital pathology. However, deployment in disease-specific settings is limited by 1) the high computational cost of billion-parameter PFMs and 2) distribution mismatch and non-biological bias inherited from pan-cancer, multi-centre pre-training, including site-specific signatures and imbalanced disease prevalence. These factors can encourage shortcut learning and under-emphasise subtle morphology required for reliable modelling of a specific cancer type. We present SmartStu (a Smart Student), a framework to customise compact, breast-cancer-specific PFMs via distillation whilst mitigating confounding. SmartStu distils representations from multiple teacher PFMs into a lightweight student backbone. Crucially, we introduce adversarial distillation that leverages a dedicated noise model trained to predict nuisance, edge-dominated cues on the distillation set. Using this noise model as a counterexample, the adversarial objective encourages the student to recognise, yet suppress, features predictive of nuisance targets. We further incorporate multi-teacher ensemble distillation and an auxiliary self-supervised objective with artefact injection. We validate SmartStu on three external cohorts (Yale HER2, SLN-Breast, and BRACS) with multiple tiny backbones. SmartStu yields breast-cancer-specific PFMs that are over $30\times$ smaller than general PFMs whilst largely preserving, and sometimes improving, downstream performance measured by balanced accuracy (bAcc) and AUC. Code is available at https://github.com/zwchen03/advDistall.
Pathology foundation models (PFMs) provide strong tile-level representations via self-supervised pre-training on large-scale pathology images. Yet, PFMs are developed under diverse and often opaque data, architecture, and objective choices, inducing latent representational biases that limit robustness and obscure what each model specialises in. We present AdaFusion, a lightweight adaptive fusion framework that integrates complementary signals from multiple frozen PFMs through (1) low-dimensional feature compression and (2) a sample-conditioned gating module that reweights model-wise (and optionally channel-wise) contributions. Beyond improving predictive accuracy, AdaFusion provides contribution-driven interpretation that offers evidence consistent with model-specific preferences and synergistic interactions across tissue phenotypes. We evaluate AdaFusion on three public benchmarks spanning treatment response prediction, prostate cancer grading, and spatial gene expression inference. AdaFusion consistently outperforms individual PFMs and other fusion baselines, while providing interpretable tissue visualisation which aligns model preferences with morphological patterns. Code is available at: https://github.com/xyx-98/PathoOracle.
Text-conditioned human motion generation has made rapid progress with the emergence of large-scale motion--language datasets. However, even datasets with rich long-form descriptions typically provide supervision only at the clip level, without explicit temporal correspondence between motion frames and language. This limits fine-grained motion--text grounding and temporally precise generation. We propose FineMoLA, a weakly supervised framework that learns fine-grained frame--phrase correspondence directly from clip-level annotations. Our method first segments long-form descriptions into action-bearing phrases, and then formulates motion--language alignment as an optimal transport problem, which naturally models many-to-many relations between motion frames and text under global constraints. With entropic regularization and Sinkhorn iterations, FineMoLA efficiently infers pseudo frame-level alignments without human labeling. Experiments on SnapMoGen demonstrate that the learned alignments outperform baselines in motion--text grounding.
Training-Free Out-of-Distribution Detection for Pathology Whole-Slide Images
无训练的病理全切片图像异常检测
Kahya, Sabri Mustafa, Chen, Richard R., Yavuz, Muhammet Sami, Lou, Jerry Jierui, Adeleye, Akanimoh, Kahya, Haci Ali, Lipkova, Jana
Abstract
Safe deployment of AI methods in medicine requires robust guardrails that detect when input data deviate from the training distribution to ensure that models provide predictions only within their scope of expertise and abstain otherwise. Out-of-distribution (OOD) detection can provide such safeguards and is extensively studied in general computer vision. Yet, it remains underdeveloped in computational pathology, where gigapixel whole-slide images (WSIs), subtle differences between disease subtypes, and variability in tissue preparation pose unique challenges for conventional OOD methods. We propose ZIO, a training-free, multimodal OOD detector for pathology WSIs that leverages vision--language pathology foundation models (FMs). ZIO constructs text and visual prototypes of in-distribution classes and integrates their complementary information through a prototype shrinkage mechanism to derive OOD scores. We provide the ZIO formulation for both slide- and patch-level FMs. We evaluate ZIO across diverse clinically relevant domain shifts, including rare diseases and near-OOD settings. Extensive evaluation of over 14,700 WSIs from five independent consortia shows that ZIO consistently outperforms both unimodal prototypes and 40 state-of-the-art OOD methods. These results demonstrate the benefits of multimodal representation for OOD detection and pave the way towards safer AI deployment in clinical practice.
PackingGPT: 3D Packing Agent for Real Furniture in Last-Mile Delivery
PackingGPT：用于最后一公里配送的真实家具三维装箱智能体
You, Yi, Li, Hui
Abstract
3D bin packing rectangular items into standardised containers to maximise space utilisation under geometric shipping automation. Loading a furniture purchase into a personal vehicle is the same task, but under more complex conditions that standard container loading algorithms ignore. This paper addresses the physically stable placement under these realistic conditions with heterogeneous boxes (e.g. varying dimensions and weights) and occupied containers (e.g. groceries). This paper provides a real-world benchmark dataset and baseline model for the Heterogeneous furniture-in-vehicle packing task. The dataset uses real furniture company flat-pack packaging data covering a large number of catalogue products via family-level extrapolation with diversity length, widths, heights, and weights. We also propose a PackingGPT framework for packing as a sequential placement inspired by the Lego assembly process, where heterogeneous boxes of varying dimensions (bricks) are placed step-by-step into the irregular remaining cargo space (creations). Five baseline packing methods were tested on our dataset without considering the Centre-of- Mass (CoM) constraints. In sedan car simulations, 10-40% of placed boxes failed the stability check on average. When the LLP model was trained on packing sequences with CoM constraints enforced during placement, the failure rate dropped to 0.67% (SUV-500).
Long-Horizon Embodied Decision-Making via Multimodal Memory Compression
通过多模态记忆压缩实现长视域具身决策
Li, Bingxuan, Yang, Rui, Qian, Cheng, Liu, Jiateng, Kim, Jeonghwan, Wang, Zhenhailong, Li, Manling, Zhang, Tong, Ji, Heng
Abstract
Agents are increasingly expected to act not only as task executors, but also as decision-makers on behalf of human users. This shift requires agents to accumulate evidence over long horizons, interpret implicit user preferences, and compare multiple candidates under partial observations. In this work, we propose DunphyBench, a new benchmark for evaluating agents on long-horizon human-centered embodied decision-making, where the agent must navigate through multiple embodied housing environments and make decisions that align with multi-dimensional human preferences. Unlike standard embodied reasoning tasks that often focus on procedural planning or immediate goal completion, our setting requires agents to integrate multimodal, multi-source input into coherent knowledge that supports complex reasoning across long horizon. The evaluation results reveal that there is a substantial gap between current agents and human performance. Furthermore, our diagnosis of state-of-the-art VLM-driven agents reveals that memory management is one of the bottlenecks, where raw multimodal history introduces noise that hinders decision quality. Motivated by this finding, we design MeMento, a preference-conditioned multimodal memory compressor that selectively compresses decision-relevant information from long-horizon history based on user preferences with a fixed set of memory tokens. Experiments show that MeMento helps VLM-driven agents improve accuracy by 7.18%, while reducing memory usage by 85.38% compared to the strongest baseline.
Event cameras produce sparse and asynchronous event streams that provide rich spatio-temporal information for efficient perception. Recent advances in event-based models have demonstrated strong performance by directly modeling asynchronous events without dense frame reconstruction. However, identifying the event-level evidence behind their predictions is crucial for improving model transparency and reliability. Directly adapting point-level saliency methods from point clouds provides fine-grained attribution but overlooks event-specific spatio-temporal structures. To address this limitation, we propose Voxel-Guided Global Event Ranking (VGER), a training-free attribution framework for point-based event cloud networks. VGER combines event-level gradient evidence with task-aware voxel perturbation evidence, transferring regional contribution into event-level attribution scores while preserving fine-grained resolution. Furthermore, VGER introduces a unified event ranking strategy, where high-ranked events are expected to be prediction-critical and low-ranked events are expected to have limited influence on predictions. We evaluate VGER on three event-based benchmarks with PointNet, PointNet++, and EventMamba. Across nine dataset-backbone settings, VGER consistently improves both high-tail and low-tail deletion performance over point-level saliency baselines.
Chinese Translation
事件相机产生稀疏且异步的事件流，提供丰富的时空信息以实现高效感知。最近在基于事件的模型中的进展表明，通过直接建模异步事件而无需密集帧重建，能够实现强大的性能。然而，识别其预测背后的事件级证据对于提高模型的透明度和可靠性至关重要。直接将点云中的点级显著性方法应用于事件云提供了细粒度的归因，但忽视了事件特定的时空结构。为了解决这一局限性，我们提出了体素引导全球事件排名（Voxel-Guided Global Event Ranking，VGER），这是一个无训练的归因框架，适用于基于点的事件云网络。VGER结合了事件级梯度证据与任务感知的体素扰动证据，将区域贡献转化为事件级归因分数，同时保持细粒度的分辨率。此外，VGER引入了统一的事件排名策略，其中高排名事件被期望对预测至关重要，而低排名事件则被期望对预测影响有限。我们在三个基于事件的基准上评估了VGER，使用了PointNet、PointNet++和EventMamba。在九个数据集-骨干网络设置中，VGER在点级显著性基线的高尾和低尾删除性能上始终表现出一致的提升。
Slot2Text: Object-Centric Visual Tokenization for Efficient and Spatially Traceable Surgical MLLMs
Slot2Text：面向对象的视觉标记化以实现高效且可追踪空间的外科手术多模态大语言模型
Liao, Guiqiu, Jogan, Matjaz, Hashimoto, Daniel A.
Abstract
Multimodal large language models (MLLM) for surgical scene understanding typically inject hundreds of dense visual tokens into a language model, leading to costly inference and limited spatial traceability for generated answers. We present Slot2Text, a dual-mode surgical MLLM that replaces dense representations of visual input with a compact set of regions encoded as slot latents. Instead of relying on contrastive alignment of the visual encoder with language, Slot2Text groups self-supervised vision features into a few regions--slots that are consumed by the language model as area-labeled visual tokens. Slot2Text-Fast uses the slot prefix to answer surgical questions. Slot2Text-Reason also identifies and locates areas relevant for reasoning, linking language outputs to corresponding slot tokens, masks or regions. Experiments on multiple visual question answering and visual grounding benchmarks show that Slot2Text-Fast is competitive with state-of-the-art baseline at a much lower cost, reducing the average total token consumption by a 91.8\% and the visual prefix from 1,295 to 47 tokens (a 96.4\% reduction). Slot2Text-Reason trades additional tokens and latency for explicit area identities, locations, and traceable spatial evidence. These results establish compact slot latents as an efficient default visual interface for surgical MLLMs, with grounded reasoning invoked when greater spatial traceability is required.
Unified multimodal object tracking has achieved remarkable robustness by leveraging complementary sensor data (e.g., RGB, Thermal, Depth), yet the heavy computational burden of state-of-the-art models hinders their deployment on resource-constrained edge devices. In this work, we identify the prediction head as a critical but often overlooked efficiency bottleneck. By strategically streamlining the decoder architecture, we unlock the potential for real-time inference but simultaneously introduce a capacity gap between the lightweight student and the heavy teacher. To resolve this, we conduct a systematic analysis of 17 distillation strategies and introduce a Dual-Alignment Distillation framework. Our key insight is that effective compression requires decoupling knowledge transfer into two complementary streams: (1) Spatial Representation Alignment, which employs feature distillation to sharpen the student's spatial focus on foreground targets ("Where to track"); and (2) Semantic Distribution Alignment, which utilizes logit-based distillation to align decision boundaries and transfer discriminative dark knowledge ("What to track"). Extensive experiments across five benchmarks demonstrate that our approach significantly outperforms complex state-of-the-art methods. Notably, our distilled model achieves 91.5% MPR on RGBT234 and operates at 54 FPS on a single RTX 4090, representing a 5x speedup over the teacher model while maintaining superior accuracy.
Selecting a complete 3D object from a reconstructed scene with minimal user effort is essential for practical scene editing and embodied interaction. Existing 3DGS-based methods either retrain the Gaussian representation to embed per-object labels, or build dense multi-view SAM observations, both requiring heavy computation and dense viewpoint coverage that is rarely available in practice. We present GaussianSelector, a training-free framework for interactive 3D object selection from sparse views and sparse scribble guidance. Operating directly on native Gaussian primitives, we coarsen dense Gaussians into geometrically coherent superpoints and construct a continuity-weighted graph using appearance and spatial cues. Sparse user scribbles are lifted into 3D via visibility-aware transmittance coverage, and selection is solved as a global graph-cut energy minimization that propagates sparse evidence to a complete 3D object. This design naturally supports multi-round refinement, where users iteratively correct the selection from additional viewpoints to progressively improve the result. Experiments demonstrate that GaussianSelector achieves competitive selection quality against state-of-the-art multi-view SAM-based methods, while requiring significantly fewer interaction views and substantially lower computational overhead. These properties make it well suited for human-in-the-loop 3D scene editing and 3D asset extraction in real-world deployment scenarios.
Probing the 3D Object-Level Understanding of Pre-Trained Detection Transformers
探究预训练检测变换器的三维物体级理解
Kim, Robin, Samplawski, Colin, Marlin, Benjamin M.
Abstract
Detection transformer models, including DETR and its extensions, learn to output a set of object-level embeddings that can be simultaneously decoded into 2D bounding boxes and class distributions. In this paper, we investigate what pre-trained 2D detection transformers understand about the 3D properties of objects. Specifically, we investigate the extent to which properties including the depth of objects from the camera and the 3D location of objects relative to the camera can be recovered from object-level embeddings using linear and non-linear probes. Across a range of detection transformer models, our results show a surprisingly strong and previously unknown ability of 2D DETR models to represent useful information about the 3D properties of objects, despite the complete lack of 3D supervision during model pre-training.
Rolling shutter (RS) cameras are widely used in consumer devices, but their row-wise exposure causes distortions under motion, making geometric 3D vision problems dependent on both camera intrinsics and readout time ratio. Existing RS calibration methods rely on calibration targets or specialised hardware, limiting their use in unconstrained settings. We present the first self-calibration method for RS cameras that directly estimates camera intrinsics and the readout time ratio from image sequences, without requiring calibration targets. The method is implemented as a self-calibrating bundle adjustment (BA), which critically depends on the RS imaging model. We combine two known complementary models. The first formulates RS imaging as continuous-time trajectory estimation under a row-wise pose representation. The second interprets RS images as temporally distorted global shutter (GS) images and requires to estimate correction fields. The combination is non-trivial and results in a unified dual-projection model, in which each 3D point is simultaneously constrained at both row-dependent and reference timestamps along a shared continuous trajectory, enforcing stronger geometric and temporal consistency. Extensive simulations analyse the applicability of several implementations under varying conditions, and real data experiments demonstrate the accuracy, robustness, and practical effectiveness of the proposed approach.
UCBound-Net: Uncertainty-Guided Boundary-Aware Continual Learning for Domain-Incremental Ultrasound Segmentation
UCBound-Net：基于不确定性引导的边界感知持续学习用于领域增量超声分割
Rahman, Mohammad Amanour
Abstract
Continual learning in clinical imaging faces a dual challenge: a model must assimilate knowledge from new anatomical domains while retaining representations learned from prior tasks, a problem known as catastrophic forgetting. Existing mitigation strategies, including regularization and knowledge distillation, treat all spatial regions equally, ignoring the fact that prediction uncertainty is strongly correlated with the propensity for forgetting. We introduce UCBound-Net, a continual segmentation framework that exploits Monte Carlo (MC) Dropout uncertainty as a spatial proxy for forgetting risk. Our method contributes three synergistic components: (i) uncertainty-weighted boundary distillation, which amplifies the knowledge transfer signal at high-entropy regions of the frozen teacher; (ii) uncertainty-calibration regularization, which explicitly penalizes overconfident erroneous predictions; and (iii) uncertainty-guided exemplar selection, a memory buffer that preferentially stores samples whose boundary regions exhibit the highest predictive entropy. Evaluated on a sequential domain-incremental benchmark comprising breast ultrasound (BUSI, Task 1) followed by thyroid ultrasound (TN3K, Task 2), UCBound-Net reduces forgetting relative to naive fine-tuning, achieving a backward transfer (BWT) of -0.098 compared with -0.173, while obtaining an average Dice Similarity Coefficient (DSC) of 0.755 across both tasks. The proposed framework outperforms baseline methods without requiring task-boundary supervision. An ablation study further demonstrates that each component contributes independently to forgetting mitigation, providing a practical pathway toward uncertainty-aware continual learning for clinical image segmentation.
Reliable decision-support in digital agriculture requires accurate predictions and well-calibrated uncertainty estimates, particularly for dense prediction tasks such as semantic segmentation. Ensemble methods provide strong uncertainty quantification, but their computational and memory demands limit practical use, while single-model approximations often trade off uncertainty quality for efficiency. We propose ST-LoRA, a parameter-efficient ensemble framework that builds diverse ensemble members from a single training trajectory by combining Low-Rank Adaptation (LoRA) with snapshot ensembling. Each member shares a frozen pretrained backbone and differs only in lightweight low-rank adapters, reducing trainable parameters to under 10% of the full model while preserving ensemble diversity. We evaluate across two agricultural datasets - GrowliFlower-L (cauliflower, open field) and BUP20 (sweet pepper, glasshouse) - using SegFormer and Mask2Former, covering in-distribution performance, calibration under distribution shift, and out-of-distribution detection. Ablations show feed-forward layers, not attention layers, are the critical LoRA target for dense prediction, contrary to the attention-only convention from language models. ST-LoRA matches or exceeds full-rank ensembles in segmentation accuracy and calibration across both datasets and architectures, while substantially reducing training time, inference latency, memory footprint, and storage requirements. Against efficient baselines - Snapshot Ensemble, MC Dropout, and Deep Deterministic Uncertainty - ST-LoRA consistently matches or outperforms them in image/pixel-level OoD detection, calibration stability under shift, and cross-seed variance, with far fewer parameters and lower compute. These results show LoRA-efficient ensemble adaptation is a highly effective, practical approach for uncertainty-aware agricultural vision systems.
Hard symbolic-reasoning tasks such as Sudoku, maze pathfinding, and ARC remain challenging for LLMs due to their fixed-depth autoregressive reasoning, which limits systematic search, refinement, and backtracking. While recursive models such as Hierarchical Reasoning Model (HRM) and Tiny Recursive Model (TRM) address this limitation through iterative latent-state refinement, they are typically task-specific and do not leverage pretrained language priors. We propose R-Qwen, a recursive reasoning framework built upon a pretrained Qwen backbone. R-Qwen repeatedly refines a candidate solution through programmatic self-recursion and deep supervision, combining the structured iterative computation of recursive models with the linguistic and reasoning priors of pretrained LLMs. We further adapt Hierarchical Supervision Weighting (HSW) to autoregressive models by exponentially weighting losses across recursive steps. HSW reduces gradient variance by at least 50\%, improves the signal-to-noise ratio of stochastic gradients, and accelerates convergence. Across eight challenging benchmarks, R-Qwen consistently outperforms prior recursive reasoning models and substantially larger LLMs while using a comparable number of trainable parameters. Notably, on ARC-AGI dataset, our model achieves a 27.6\% improvement over the baseline, highlighting the effectiveness of recursive refinement for general symbolic reasoning. These results suggest that recursive reasoning mechanisms and pretrained language model priors are complementary approaches for improving symbolic puzzle-solving. Code and models will be released after acceptance.
Vision-language models (VLMs) are emerging as a key component of embodied intelligence, with growing applications in auto-labeling and end-to-end autonomous driving. However, existing approaches for improving spatiotemporal reasoning in VLMs often rely on complex preprocessing pipelines, expensive human annotations, or synthetic data, which limit scalability and introduce potential sim-to-real gaps. Moreover, although these methods have improved spatiotemporal understanding, they still lack strong metric reasoning capabilities for dynamic scenes, such as estimating object motion in real-world units. Prior work has explored LiDAR-based metric depth supervision to enhance spatial perception, but it does not directly address temporal reasoning. We introduce STAR-VLM, an automotive radar-supervised framework that enhances spatiotemporal VLMs with motion reasoning and metric velocity estimation for autonomous driving. Automotive radar is a low-cost and widely deployed sensor that provides complementary spatiotemporal supervision through range and Doppler measurements. By leveraging these measurements as label-free ground truth during training, STAR-VLM improves the metric spatiotemporal reasoning ability of VLMs. Through experiments on driving scenarios, we show that STAR-VLM achieves state-of-the-art performance on both motion classification and metric velocity estimation, outperforming even task-specific methods designed for each task. These results highlight automotive radar as a scalable and cost-effective source of supervision for building metric-aware spatiotemporal VLMs for real-world autonomous driving.
Enhancing Visual Perception in Foggy Conditions via Multiclass Fog Density Modeling
通过多类雾密度建模增强雾天条件下的视觉感知
Chaar, Mohamad Mofeed, Weidl, Galia
Abstract
Autonomous driving (AD) systems have advanced rapidly over the past decade; however, robust perception under adverse weather conditions remains a major challenge, particularly in dense fog. In this work, we investigate fog-aware perception using synthetically generated fog data derived from the Waymo dataset. To support fog simulation, depth images are generated using an iterative learning approach. We consider five fog-density levels: clear, light fog, moderate fog, heavy fog, and very heavy fog. Instead of training a single unified model across all conditions, we train separate perception models for each fog-density level. Experimental results show that density-specific training improves performance in severe fog conditions. In particular, for the very heavy fog class, recall improves from 0.076 to 0.232, corresponding to an absolute gain of 15.6 percentage points. These findings suggest that deploying multiple specialized models, rather than a single general-purpose model, can improve perception robustness for autonomous vehicles under challenging visibility conditions. Future work will extend this strategy to additional sensing modalities, including LiDAR and radar, and evaluate generalization across diverse weather scenarios.
Dynamic 4D Gaussian Splatting has emerged as an efficient representation for dynamic novel view synthesis through explicit scene modeling and real-time rendering. However, existing methods typically require dense multi-view videos for sufficient geometric constraints, making capture expensive and limiting sparse-camera deployment. Reducing input views lowers acquisition cost but weakens geometry supervision, often causing missing structures and floating Gaussians. Depth priors provide geometric cues, yet no single source offers both dense coverage and reliable geometry. Monocular depth provides dense structure but is scale-ambiguous and locally biased, whereas multi-view geometric depth provides incomplete anchors consistent with the reconstruction coordinate system. To exploit their complementarity, we propose D$^2$-4DGS, a sparse-camera dynamic 4D Gaussian Splatting framework guided by dual-source depth priors. We align monocular estimates with valid multi-view geometric depths and verify their consistency to identify reliable geometric anchors. These verified anchors support consistency-aware pruning and depth supervision, while verified geometric depths and aligned mono-only estimates provide candidate geometry for densification in under-reconstructed regions. Finally, RGB-D joint optimization improves appearance fidelity and geometric consistency under sparse-view supervision. Across all nine dataset--view settings, D$^2$-4DGS achieves the highest PSNR, improving by 1.33 dB on average over the best competing method in each setting.
When Measurement Conventions Masquerade as Calibration Gains in Cardiac Digital Twins
当测量约定伪装成心脏数字双胞胎中的校准增益
Cao, Dang P. M., Pham, Hieu
Abstract
Cardiac digital twins convert clinical images into physiological measurements through observation operators, yet calibration studies often assume a fixed reference convention. Across four shared-backbone echocardiographic EF front-ends, phase conditioning appears to remove CAMUS baseline bias. Matched-reference analysis rejects this gain: singleplane ground-truth EF error is statistically indistinguishable across models, while single-plane ground-truth EF exceeds CAMUS biplane clinical EF by +6.30 points, explaining nearly all baseline bias. A prespecified EchoNet-Dynamic replication, with released data and our extractor aligned to the apical four-chamber plane, removes baseline overestimation and reverses the CAMUS ranking. We also quantify haemodynamic effects, conformal residual-width budgets, and EF-stratum changes, yielding a Convention-Aware EF Audit protocol that separates genuine observation operator calibration from measurement artefacts. GitHub: EjectionFraction-Bias-in-Cardiac-Digital-Twin.git
Linear Multi-Timescale Retention as a Memory-Efficient Vision-Language Bridge
线性多时间尺度保留作为一种内存高效的视觉-语言桥梁
Yeafi, Ashfak, Hasan, Mehedi, Islam, Md Khairul
Abstract
Vision-Language Models (VLMs) face a critical computational bottleneck when processing high-resolution imagery due to the $O(N^2)$ memory complexity of Softmax Multi-Head Attention (MHA). While substituting MHA with independent Multi-Layer Perceptrons (MLPs) achieves $O(N)$ scaling, it strips the architecture of spatial sequence routing, severely degrading global scene understanding and object permanence. In this paper, we propose the Linear Multi-Timescale Retention (LIA-MTR) module, a memory-efficient cross-modal bridge. By integrating an ELU-based positive feature mapping with adaptive write-gating and log-linearly distributed recurrent decays, LIA-MTR mathematically compresses continuous visual sequences into bounded memory states. Theoretical analysis proves the architecture operates with strict $O(N)$ sequence-interaction complexity. Empirically, synthetic retrieval evaluations demonstrate that LIA-MTR flawlessly routes context across 16,000 tokens, eliminating the "Lost in the Middle" degradation typical of naive linear attention. Hardware benchmarking reveals infinite-context scaling capabilities, natively processing 262,144 visual patches within an 11.2 GB VRAM footprint, whereas standard MHA suffers out-of-memory failure at 16,384 patches. Furthermore, following instruction tuning on 665K conversational samples, LIA-MTR significantly outperforms an industry-standard MLP baseline on the MME benchmark (71.00% vs. 68.11%), driven by a 10% absolute improvement in object permanence and superior global semantic extraction. This work establishes a mathematically rigorous, computationally flat foundation for infinite-context Vision-Language integration.
Motion Beyond Morphology: Bootstrapping Cross-Category Motion Transfer from Abstract Motion Representations
超越形态的运动：从抽象运动表示中引导跨类别运动转移
Fang, Zhixue, Zhang, Zhimin, Du, Bi'an, Meng, Zijie, Zhou, Yan, Hu, Wei, Zhang, Guoxin, Wan, Pengfei, Gai, Kun
Abstract
Video motion transfer aims to animate a target object using dynamics from a reference video. Existing formulations largely rely on fixed structural correspondence, which becomes ill-defined when reference and target objects differ substantially in morphology, articulation, or deformation mechanisms. We introduce Motion Beyond Morphology, a perspective that seeks to transfer motion beyond fixed structural correspondence, by preserving dynamics that remain meaningful across different target morphologies. To realize this, we propose a two-stage framework. Stage~I learns complementary multi-granularity abstract motion views and uses them to bootstrap cross-category video pairs that preserve transferable dynamics across diverse morphologies. Stage~II internalizes this supervision into direct reference-video-conditioned generation, removing the need for explicit motion extraction at inference. We further introduce OpenVMT-Dataset and OpenVMT-Bench for training and evaluating image- and text-conditioned motion transfer across Same, Near, and Far category gaps, and plan to release both upon acceptance. Extensive experiments demonstrate state-of-the-art motion fidelity and target preservation. Project page: https://miniz233.github.io/MotionBeyondMorphology/
Despite the progress of multimodal large language models (MLLMs), they continue to exhibit deficiencies in visual perception. Following visual instruction tuning, internal MLLM representations rapidly deviate from their original semantic states during inference, causing severe information degradation. While existing methods attempt to leverage external vision foundation models (VFMs) to align internal representations, we find that direct alignment with VFMs enhances visual semantics but fails to mitigate representation deviation. To address this, we propose Spatial-Spectral Visual Anchor Learning (SSVAL). The core of SSVAL is Visual Anchor Prompt Injection (VAPI), which introduces prompts that absorb rich knowledge from external VFMs during training, enabling them to serve as stable visual anchors that mitigate representation deviation during inference. Additionally, we incorporate auxiliary spatial and frequency-domain representation alignment losses to provide complementary vision-specific supervision at intermediate LLM layers. Extensive experiments demonstrate that SSVAL significantly outperforms existing methods. Code are available on our \href{https://msls38.github.io/SSVAL/}{project page}.
Egocentric visual grounding requires high-resolution inputs to localize small objects. However, scaling Multimodal Large Language Models to this domain is constrained by the excessive cost of visual token processing. We identify that current efficient strategies based on token reduction are unreliable for selecting object-centric spatial evidence. To overcome this, we propose SmartRes, a framework that performs efficiency optimization in the pixel space via dynamic resolution routing. SmartRes first encodes a low-resolution view for global context and uses a lightweight router to activate high-resolution patches in object-centric regions and constructs an order-preserving visual sequence. To further enable robust routing under severe foreground-background imbalance, we introduce a margin-regularized routing objective that increases foreground-background logit separation and improves foreground recall. Experiments on Ego4D and EgoIntention show that SmartRes reduces visual tokens by up to 67% while retaining 86.4% of full-resolution performance, and achieves up to 1.66X faster inference than state-of-the-art token reduction methods with higher accuracy. Furthermore, strong performance on small object grounding indicates the effectiveness of SmartRes towards egocentric applications. Code will be publicly available.
StreamTalk: Streaming Co-Speech Gesture Generation with Key-Pose Anchoring
StreamTalk：基于关键姿态锚定的流式共语手势生成
Zhang, Xiangyue, Li, Jianfang, Zhang, Jiaxu, Yang, Kaixing, Hoi, Steven
Abstract
Real-time co-speech gesture generation must produce 3D motion clip by clip as speech arrives. Existing streaming methods are open-loop: each clip depends on past context, but the model cannot check or correct its trajectory. Small errors therefore accumulate and cause drift over long sequences. We observe that this failure is mainly caused by the lack of a forward constraint rather than poor short-clip quality. A plausible key pose at the end of each clip provides a destination anchor that limits drift. Based on this observation, we propose StreamTalk, a closed-loop framework with a periodic generate-retrieve-refine cycle. Streaming Pose-Guided Generation first predicts a coarse clip, retrieves a plausible tail pose from a speaker-specific motion database, and refines the clip using this pose before continuing to the next window. During training, Stochastic Anchor Masking randomly masks pose and translation frames, teaching the model to recover complete motion from sparse boundary conditions. A part-aware DiT separates hand, body, and translation streams to reduce interference between global displacement and local articulation. On BEAT2, StreamTalk achieves state-of-the-art FGD, reduces long-horizon drift relative to open-loop baselines, and runs in real time at 76 FPS. Project page: https://xiangyue-zhang.github.io/StreamTalk/.
In video understanding, vision-language models (VLMs) must ingest massive numbers of visual tokens, causing the computational and memory cost of the prefill stage to rise sharply. Such visual sequences are highly redundant along the spatio-temporal dimension, yet a high compression ratio is often accompanied by the loss of critical details. Existing token-compression methods either employ heuristic, training-free compression with limited content adaptivity or introduce additional modules that require expensive alignment training, leaving the trade-off between efficiency and adaptivity unresolved. To alleviate this limitation, we propose CRAFT: Compression via Recursive Adaptive Fusion of Video Tokens. CRAFT recursively merges tokens by decoupling parameter-free token selection from learnable token fusion: global similarity determines which tokens to merge, while a position-aware weighting module and a content-adaptive channel-wise gate learn how to fuse them. The whole compression pipeline is query-agnostic. Because every retained token is a linear combination of the original tokens, CRAFT preserves their true spatio-temporal coordinates and stays aligned with the pre-trained language model's input distribution. Experiments on multiple representative video benchmarks show that CRAFT consistently outperforms prior state-of-the-art token-compression methods. At about $8\times$ compression, it retains roughly $97\%$ of the backbone's average accuracy and shows significant efficiency improvement.
Feed-forward 3D Gaussian Splatting enables efficient novel-view synthesis without per-scene optimization, but most existing methods assume a fixed set of context views and process them jointly. This limits their applicability to online scenarios where calibrated views arrive sequentially and the scene must be updated causally. We present \emph{StreamSplat}, a streaming feed-forward 3DGS framework that incrementally maintains a persistent geometry-grounded scene state and decodes it into renderable 3D Gaussians after each input chunk. StreamSplat centers on a \textbf{Voxel-Aligned Causal Cache (VACC)}, which stores historical 3D tokens in a memory-bounded voxel structure so that memory grows with explored scene geometry rather than stream length. To better reuse history during causal prediction, we introduce \textbf{History-Projected Depth Anchoring (HPDA)} to project cached geometry as depth guidance for current cost-volume estimation, and \textbf{Cache-Guided Feature Injection (CGFI)} to inject cached latent evidence into Gaussian-token regression. Experiments on DL3DV, RealEstate10K, and ScanNet show that StreamSplat remains competitive with state-of-the-art feed-forward 3DGS methods under sparse causal inputs, despite not using future views or full-scene context. More importantly, it scales to long input streams with 256, 512, and 1024 views where fixed-view baselines run out of memory, yielding sustained improvements in novel-view synthesis quality as more observations arrive. The code will be made publicly available upon acceptance.
Long-video question answering requires identifying sparse yet critical evidence from videos containing thousands of frames under a constrained visual-token budget. Existing methods either select query-aware frames in a single pass or rely on timestamped text solely as retrieval guidance, leading to two key limitations. First, selected frames tend to cluster around local relevance peaks, and once the budget is exhausted, omitted evidence cannot be recovered. Second, textual and visual evidence remain weakly aligned. We propose GCR, a training-free framework that casts fixed-budget frame selection as a joint evidence curation problem. Ground converts timestamped text into temporal events, selects query-relevant real frame anchors, and renders each event text onto its temporally aligned frame. Cover supplements grounded events with direct visual anchors for complementary visual evidence and applies global maximal marginal relevance to preserve diverse context. Refine revisits omitted temporal regions and replaces the weakest revisable context frame with a real-frame medoid---but only when the medoid offers greater evidence value. GCR maintains a fixed number of chronologically ordered frames and requires no VLM training or architectural modification. Experiments on LongVideoBench and Video-MME, across three 7B backbones and frame budgets of 8, 32, and 64, demonstrate consistent improvements in long-video QA. With the 7B LLaVA-OV backbone and 32 frames, GCR achieves 64.25% and 62.15% on the two benchmarks, outperforming the strongest reproduced baselines by 2.54 and 1.93 percentage points, respectively.
Chinese Translation
长视频问答需要在受限的视觉标记预算下，从包含数千帧的视频中识别稀疏但关键的证据。现有方法要么在单次处理过程中选择与查询相关的帧，要么仅依赖带时间戳的文本作为检索指导，这导致了两个主要局限。首先，所选帧往往聚集在局部相关性峰值附近，一旦预算耗尽，遗漏的证据无法恢复。其次，文本和视觉证据之间的对齐较弱。我们提出了GCR（Ground, Cover, and Refine），一个无训练的框架，将固定预算的帧选择视为一个联合证据策划问题。Ground将带时间戳的文本转换为时间事件，选择与查询相关的真实帧锚，并将每个事件文本呈现到其时间对齐的帧上。Cover通过直接的视觉锚补充已定位的事件，以提供互补的视觉证据，并应用全局最大边际相关性以保持多样化的上下文。Refine重新审视遗漏的时间区域，并用真实帧的质心替换最弱的可修订上下文帧——但仅当质心提供更大的证据价值时。GCR保持固定数量的按时间顺序排列的帧，并且不需要VLM（视觉语言模型）训练或架构修改。在LongVideoBench和Video-MME上的实验，涵盖三个7B骨干网络和8、32、64的帧预算，展示了长视频问答的一致性改进。在使用7B LLaVA-OV骨干网络和32帧时，GCR在两个基准上分别达到了64.25%和62.15%的成绩，分别比最强的复现基线提高了2.54和1.93个百分点。
FairForensics: Seeing Expressions and Parsing Demographics via Vision-Language Modeling for Generalizable Fair Deepfake Detection
公平取证：通过视觉-语言建模观察表情和解析人口统计信息以实现可泛化的公平深度伪造检测
Zhang, Yaning, Wu, Jiao, Gao, Zan, Shen, Linlin
Abstract
The challenge of fair deepfake detection (FDD) has attracted increasing attention. Existing fairness-enhanced detectors often suffer from suboptimal generalization to unseen manipulations and fairness across demographic groups. They are typically developed and evaluated on demographically imbalanced distributions, resulting in biased predictions toward minority groups. In this paper, we construct a novel demographically balanced FDD benchmark to train and evaluate the fairness of detectors under both balanced and imbalanced population scenarios. Additionally, we introduce a novel expression and demographic perceptual vision-language model, termed FairForensics, for generalizable fair deepfake detection. FairForensics conducts face forgery generalization enhancement and demographic-aware fairness regularization. During face forgery generalization enhancement, built upon the novel observation of significant distribution differences between pristine and forged expression vectors, we design an expression encoder to capture high-level expression-guided forgery patterns, and an expression-perceptual visual encoder that integrates global appearance and expression forgery features while mitigating identity bias using an identity-aware patch perturbation module. Under demographic-aware fairness regularization, we propose a demographic-guided language encoder to extract population-aware global language embeddings, which boosts the decoupling of forgery features from demographic information via vision-language alignment. We devise a population-aware prototype fairness objective to enforce both inter-class separability and intra-class alignment across demographic subgroups. Extensive experiments on our balanced demographic benchmark show that our method achieves the state-of-the-art in terms of generalization and fairness.
Few-Shot Concept Prompt Learning for Segmentation Foundation Models via Visual Grounding
通过视觉定位进行少样本概念提示学习的分割基础模型
Venkataramani, Rahul, Sathish, Rachana
Abstract
Promptable segmentation foundation models (FMs) such as SAM3 and Medical SAM3 promise few-shot, interactively-specified segmentation for medical imaging through a natural language interface, yet their performance on clinical tasks falls well short of this promise. We posit that this shortfall is not an artefact of insufficient medical pretraining or imperfect prompt phrasing, but a structural limitation that will persist in any domain where paired image-text supervision is scarce, as it is across most clinical modalities. We further hypothesize that the limitation is specific to natural language as a control signal: a visually grounded prompt, learned directly from the target distribution, should recover the lost performance without additional image-text data or backbone retraining. We propose Few-Shot Concept Prompt Learning (FS-CPL), which learns a continuous concept prompt embedding $\mathbf{p}^* \in \mathbb{R}^{T \times d}$ from a small support set of $K$ image--mask pairs via mask supervision, with the encoder-decoder backbone frozen. Across four public benchmarks spanning ultrasound and endoscopy (BUSI, HC18, TN3K, CVC-Clinic), FS-CPL delivers absolute Dice improvements of up to $+0.62$ over canonical text prompts and is \emph{backbone-agnostic}: it lifts both vanilla SAM3 and the domain-specifically pretrained Medical SAM3, showing that visual concept prompting is complementary to in-domain pretraining.
Chinese Translation
可提示的分割基础模型（FMs），如SAM3和Medical SAM3，通过自然语言接口承诺实现医学影像的少样本、交互指定分割，但它们在临床任务上的表现远未达到这一承诺。我们认为，这一不足并非由于医学预训练不足或提示措辞不当，而是一种结构性限制，这种限制将在任何图像-文本配对监督稀缺的领域中持续存在，正如大多数临床模式所示。我们进一步假设，这一限制特定于自然语言作为控制信号：一种直接从目标分布中学习的视觉定位提示，应该能够在没有额外图像-文本数据或主干网络重新训练的情况下恢复丢失的性能。我们提出了少样本概念提示学习（FS-CPL），它通过掩膜监督从小型支持集中的$K$个图像-掩膜对学习一个连续的概念提示嵌入$oldsymbol{p}^* ext{ in } ext{R}^{T imes d}$，同时保持编码器-解码器主干网络不变。在涵盖超声和内窥镜的四个公共基准测试（BUSI、HC18、TN3K、CVC-Clinic）中，FS-CPL在经典文本提示上提供了高达$+0.62$的绝对Dice改进，并且是 extit{主干无关}的：它提升了普通的SAM3和领域特定预训练的Medical SAM3，表明视觉概念提示与领域内预训练是互补的。
FAU at ImageCLEF 2026 Task on Multimodal Reasoning Robust Candidate Scoring and Concise Multilingual Visual Answering
FAU在ImageCLEF 2026多模态推理鲁棒候选评分与简洁多语言视觉问答任务中的表现
Basem, Mohamed, Christlein, Vincent
Abstract
We present our ImageCLEF 2026 Multimodal Reasoning system for the Visual Multiple Choice Question Answering (Visual MCQ) and Visual Open Question Answering (Visual OpenQA) subtasks. The challenge requires reliable reasoning over multilingual educational and scientific images with dense text, diagrams, charts, tables, formulas, and units, while enforcing strict answer formats. Our central finding is that robust output control is as important as model choice. For Visual MCQ, we replace fragile free-form generation with direct candidate label scoring from vision-language model logits, then combine complementary runs through score fusion and voting. For Visual OpenQA, we use image enhancement, concise final answer prompting, deterministic decoding, and targeted post-processing to remove reasoning traces and formatting artifacts. Without task-specific model training, our official submissions achieved third place in Visual MCQ with 0.7108 accuracy and first place in Visual OpenQA with 0.6488 COMET, 0.1391 BLEU, 0.2762 ROUGE L, and 0.2383 METEOR. The results highlight the practical value of inference engineering: careful scoring, ensembling, prompting, and cleanup can turn strong VLMs into reliable competition systems.
Generative Brownian Bridge Diffusion In Motion Space For Enhanced Myocardial Strain Analysis
运动空间中的生成布朗桥扩散用于增强心肌应变分析
Paul, Rishov, Epstein, Frederick H., Zhang, Miaomiao
Abstract
Myocardial strain analysis of cardiac magnetic resonance (CMR) images provides an important tool for evaluating cardiac function. However, current techniques require either human-adjusted post-processing with suboptimal regional accuracy, or specialized and costly imaging acquisitions. In this paper, we propose to leverage the power of generative models to synthesize high-quality motion-derived strain values from routinely acquired CMR sequences. Specifically, we develop a novel Brownian bridge diffusion model in motion space to learn the probabilistic mapping between standard CMR motion estimated from widely adopted registration methods and highly accurate motion provided by advanced strain imaging techniques. To promote the fidelity of anatomical structure in the generation process, our model is conditioned on the corresponding CMR images. We validate our method on large-scale multi-center CMR datasets including subjects of paired standard cine CMR and advanced strain imaging acquisitions. Experimental results demonstrate that our framework significantly improves the accuracy of motion prediction and strain analysis from standard CMRs compared to existing learning-based approaches. Our research represents a new paradigm for potentially developing cost-effective, clinically deployable AI tools for cardiac function assessment with enhanced strain accuracy in busy clinical workflows. Our code is publicly available at anonymous.4open.science/r/Brownian-Bridge-strain-analysis-1140.
Generative AI and Foundation Models in Medical Image
医疗图像中的生成性人工智能与基础模型
Oda, Masahiro
Abstract
In recent years, generative AI has attracted significant public attention, and its use has been rapidly expanding across a wide range of domains. From creative tasks such as text summarization, idea generation, and source code generation, to the streamlining of medical support tasks like diagnostic report generation and summarization, AI is now deeply involved in many areas. Today's breadth of AI applications is clearly distinct from what was seen before generative AI gained widespread recognition. Representative generative AI services include DALL-E 3 (OpenAI, California, USA) and Stable Diffusion (Stability AI, London, England, UK) for image generation, ChatGPT (OpenAI, California, USA), and Gemini (Google, California, USA) for text generation. The rise of generative AI has been influenced by advances in deep learning models and the scaling up of data, models, and computational resources based on the scaling laws. Moreover, the emergence of foundation models, which are trained on large-scale datasets and possess general-purpose knowledge applicable to various downstream tasks, is creating a new paradigm in AI development. These shifts brought about by generative AI and foundation models also profoundly impact medical image processing, fundamentally changing the framework for AI development in healthcare. This paper provides an overview of diffusion models used in image generation AI and large language models (LLMs) used in text generation AI, and introduces their applications in medical support. This paper also discusses foundation models, which are gaining attention alongside generative AI, including their construction methods and applications in the medical field. Finally, the paper explores how to develop foundation models and high-performance AI for medical support by fully utilizing national data and computational resources.
Entity-Aware Sequence Transduction for Player-Centric Ball Action Spotting
面向实体的序列转导用于以球员为中心的球类动作检测
Wang, Ruifeng, Yang, Di, Wang, Jiangtao
Abstract
Player-centric ball action spotting requires temporally precise event detection together with actor attribution in crowded, partially observed multi-agent sports videos. Existing Denoising Sequence Transduction (DST) baselines treat the player-role dimension as part of a flattened frame-level representation, which weakens the inductive bias for modeling player-specific temporal evolution and inter-player interactions. To address this limitation, we propose Multi-Entity Denoising Sequence Transduction (ME-DST). ME-DST keeps the role-slot dimension throughout encoding. It uses temporal attention to model the history of each role slot, and spatial attention to exchange information across role slots at each frame. This factorized design gives the model a direct structure for separating within-player evolution from inter-player context. We also add learnable role embeddings, tracking-derived tactical features, and fused visual predictions from X3D-L and Swin3D-S. Experiments on the FOOTPASS dataset show that ME-DST reaches a Micro F1 of 0.778. This improves the strongest official TAAD+DST baseline by 10.3 percentage points. Controlled ablations show that preserving the entity axis and encoding role identity are central to this gain. These results suggest that explicit entity modeling is an effective inductive bias for player-centric sports event understanding.
Recent geometric foundation models enable feed-forward inference for SLAM, but their predictions are strongly dependent on the input view set, which leads to geometric inconsistencies and trajectory drift when results are chained over long sequences. Online deployment further exposes a trade-off between the low latency of two-view tracking and the constraint richness of multi-view inference. We introduce UniSim-SLAM, an integrated system that runs lightweight two-view keyframe tracking in the frontend and performs periodic multi-view submap refinement in the backend. To combine predictions defined in heterogeneous local coordinates with inconsistent scales, we formulate a unified multi-level factor graph on $Sim(3)$ that jointly optimizes global keyframe poses and submap poses. The graph integrates temporal view-to-view odometry edges, view-to-submap bridge edges with depth-statistics scale anchoring, and submap-to-submap tie and scale constraints to enforce consistent similarity relations across submaps. Experiments on TUM RGB-D and 7-Scenes show that UniSim-SLAM achieves state-of-the-art accuracy in the uncalibrated setting, reducing trajectory error by $38.5\% $ on TUM RGB-D and $45.9\%$ on 7-Scenes compared to prior best results. Project page: https://vision3d-lab.github.io/unisim-slam/
SpatialQuery: Benchmarking Geometry-Grounded Multi-Instance Spatial Reasoning in Vision-Language Models
SpatialQuery：在视觉语言模型中基于几何的多实例空间推理基准测试
Nguyen, Hai, Vu, Tung, Tran, Cong
Abstract
Vision-language models (VLMs) achieve strong semantic understanding but remain unreliable in metric spatial reasoning, particularly when queries require comparing multiple instances of the same object category. We study this problem through the Closest-Instance Distance Query (CIDQ), where a model must identify the nearest visible candidate to a unique reference object and estimate their gravity-aligned floor-plane distance. We introduce SPATIALQUERY, a training- free framework for CIDQ reasoning from a single RGB image, together with SPATIALQUERY-1M, a benchmark containing over one million RGB-only question-answer pairs from 200 indoor scenes. SPATIALQUERY recovers instance-level metric geometry and transforms it into a canonical Bird's-Eye View through Scene Cubifying, which represents objects as uniformly sized, category-coded blocks to emphasize their relative floor- plane locations. We further propose Uncertainty-Aware Chain-of-Thought (UA-CoT) prompting, which incorporates geometry- derived per-instance uncertainty into the VLM reasoning process. Without task-specific fine-tuning or architectural modification, SPATIALQUERY with Qwen3-VL-8B achieves a Floor-MAE of 0.259 m, an [email protected] m of 90.5%, and a proximity-decision accuracy of 84.18%, outperforming fine-tuned spatial specialists, general-purpose VLMs, and closed-source frontier models. Code, benchmark resources, and an interactive demo are available at https://namhai1810.github.io/SpatialQuery/.
Accurate crack assessment in electroluminescence (EL) images is important for photovoltaic (PV) reliability analysis, yet existing segmentation methods often fail to capture the thin, elongated, and structurally constrained nature of crack defects. This paper proposes a Solar Topology Crack Network (STC-Net) that incorporates edge priors, spectral priors, and a boundary-topology refinement module to improve crack continuity and boundary preservation. The framework further extends segmentation to power-loss estimation by deriving a crack-associated inactive-area proxy from the predicted masks. Experiments on the PVEL-S dataset show that STC-Net achieves 95.98 MIoU, 98.01 MDice, and 98.00 MAcc during training, and 72.52 MIoU and 80.16 MDice on unseen test samples. These results demonstrate that STC-Net provides accurate crack localization while offering a practical link between EL-based defect segmentation and PV degradation assessment.
Extremely low-light RAW enhancement aims to recover severely attenuated sensor signals, yet existing methods often focus on illumination and noise while overlooking the motion-induced degradations inherent in practical low-light imaging. We present a framework for robust extremely low-light RAW enhancement under realistic acquisition degradations. First, we introduce See in the Degraded Extremely Dark (SIDED), a new dataset that applies controlled motion degradation to extremely low-light RAW pairs while retaining their original sensor noise. Second, we propose a unified RAW tokenizer equipped with explicit domain-conditioned representation calibration to align extremely low-light and well-exposed RAW data, followed by a MeanFlow that performs enhancement in a single function evaluation. To our knowledge, this is the first work to formulate extremely low-light RAW enhancement under realistic motion-degraded acquisition and address it with MeanFlow. We further introduce a physics-guided refinement model to strengthen illumination--reflectance consistency, pixel fidelity, and color preservation without incurring additional inference cost. Extensive experiments demonstrate that our framework achieves state-of-the-art performance in extremely low-light RAW enhancement, and robustly handles coupled motion and noise degradations.
Chinese Translation
极低光照RAW增强旨在恢复严重衰减的传感器信号，但现有方法往往侧重于照明和噪声，而忽视了实际低光照成像中固有的运动引起的退化。我们提出了一个在现实采集退化下进行稳健极低光照RAW增强的框架。首先，我们引入了“在退化的极端黑暗中观察”（See in the Degraded Extremely Dark, SIDED），这是一个新的数据集，应用了受控的运动退化于极低光照RAW对，同时保留了其原始传感器噪声。其次，我们提出了一种统一的RAW分词器，配备显式的领域条件表示校准，以对齐极低光照和良好曝光的RAW数据，随后采用MeanFlow在单次函数评估中进行增强。据我们所知，这是首个在现实运动退化采集下构建极低光照RAW增强的工作，并通过MeanFlow进行处理。我们进一步引入了一种物理引导的精细化模型，以增强照明-反射一致性、像素保真度和色彩保留，而无需增加额外的推理成本。大量实验表明，我们的框架在极低光照RAW增强中实现了最先进的性能，并稳健地处理了耦合的运动和噪声退化。
3D Gaussian Splatting has achieved remarkable success in photorealistic and efficient rendering, leading to a rapid increase in 3D assets represented by 3D Gaussian primitives. Directly rigging these assets with arbitrary skeleton topologies is highly desirable. However, training a feed-forward skinning framework is infeasible due to the lack of high-quality 3D Gaussian rigging datasets. An alternative solution is to transfer mesh-based techniques to 3D Gaussian-based representation, but 3D Gaussian primitives are not restricted to the surface and lack explicit topological connectivity. Moreover, this kind of method suffers from poor generalization to unseen data due to its strong dependence on training data, while acquiring high-quality rigging data is prohibitively expensive. To address this challenging problem, we propose G-Skin, a novel generative skinning framework designed for expressive and high-fidelity animation with 3D Gaussian representation. To overcome this 3D data scarcity, we introduce a skeleton-controllable image generation model leveraging 2D vision foundation models to distill powerful motion priors into pseudo-guidance. Guided by these priors, we formulate an optimization pipeline incorporating geometry-aware regularizations, which stabilizes the learning process and ensures smooth, structurally coherent skinning weights. G-Skin also generalizes flexibly to the augmented variants of 3D Gaussian representation designed to mitigate animation-induced rendering artifacts. Extensive experiments validate the effectiveness of our approach, demonstrating clear advantages over state-of-the-art methods. Project page: https://yaoyx689.github.io/GSkin.html.
Learning Where to Look and How to Judge: Resolution-agnostic Image Quality Assessment with Quality-aware Saliency
学习关注何处以及如何评判：一种与分辨率无关的图像质量评估方法，结合质量感知显著性
Gedik, Hakan Emre, Gupta, Shashank, Bovik, Alan
Abstract
No-reference image quality assessment (NR IQA) has recently benefited from deep and multimodal models, yet many SOTA systems still violate at least one basic requirement: they either discard critical quality cues via aggressive resizing, fail to generalize across resolutions, cannot be jointly trained on heterogeneous IQA datasets with mismatched MOS scales, or require prohibitive computation. We present \textbf{ReLIQS}, a model for \textbf{Re}solution-agnostic \textbf{L}earning for \textbf{I}mage \textbf{Q}uality with \textbf{S}aliency, which is resolution-agnostic, preserves original-resolution quality cues, learns from multiple subjective studies, and remains computationally efficient and budget-adaptive. ReLIQS is a CLIP-based multiscale patch-driven architecture that learns both \emph{where to look} and \emph{how to judge} quality. Fixed-size patches are sampled across multiple resolutions, including the original resolution, and encoded with a CLIP vision backbone. A lightweight Perceptual Importance Estimator then predicts IQA-specific importance maps to select a small set of informative patches, and a Latent Quality Axis Module aggregates their embeddings into a single image-level score. Across authentic, synthetic, and AIGC benchmarks spanning diverse resolutions and distortions, ReLIQS generalizes better than strong CNN-, CLIP-, and MLLM-based baselines with matching or reduced computational cost.
Chinese Translation
无参考图像质量评估（NR IQA）最近受益于深度学习和多模态模型，然而许多最先进的系统仍然违反至少一个基本要求：它们要么通过激进的缩放丢弃关键的质量线索，要么无法在不同分辨率之间进行泛化，无法在具有不匹配的主观满意度（MOS）尺度的异构IQA数据集上进行联合训练，或者需要过高的计算成本。我们提出了 extbf{ReLIQS}，一种用于图像质量的 extbf{Re}solution-agnostic extbf{L}earning with extbf{S}aliency的模型，它与分辨率无关，保留原始分辨率的质量线索，从多个主观研究中学习，并保持计算效率和预算适应性。ReLIQS是一个基于CLIP的多尺度补丁驱动架构，学习 extit{关注何处}和 extit{如何评判}质量。固定大小的补丁在多个分辨率（包括原始分辨率）中进行采样，并通过CLIP视觉主干进行编码。然后，一个轻量级的感知重要性估计器预测特定于IQA的重要性图，以选择一小组信息丰富的补丁，而潜在质量轴模块将它们的嵌入聚合为一个单一的图像级评分。在涵盖多种分辨率和失真的真实、合成和AIGC基准测试中，ReLIQS的泛化能力优于强大的基于CNN、CLIP和MLLM的基线，同时计算成本相当或降低。
IDraw: Artist Verification from Digital Drawing Images
IDraw：基于数字绘图图像的艺术家验证
Kim, Nayoung, Jiang, Nan, Sun, Bangjie, Shin, Jaewon, Kim, Sojeong, Han, Jun
Abstract
As digital drawings are increasingly shared online, reliable authorship verification has become important for protecting artists and resolving disputes. Yet when authorship is questioned, verification may have to rely only on the disputed drawing and reference drawings known to be created by the claimed artist. This setting is challenging for two reasons. First, artist-specific drawing behavior, such as pen pressure and movement speed, is informative but is not available from a completed drawing. Second, similarities in the depicted object or scene can obscure similarities arising from the artist. We propose IDraw, a framework that learns from drawings paired with tablet-pen sensor signals collected from separate training artists. This allows IDraw to infer drawing behavior from completed images during a later authorship dispute, without requiring sensor data from the artist being verified. IDraw also reduces the influence of drawing content by identifying information shared by drawings of the same object across different artists and suppressing it before comparing drawings. To support this approach, we construct the first multimodal dataset for digital drawing authorship verification, containing 1,110 drawings from 37 artists and 14 types of tablet-pen sensor signals. Evaluated on previously unseen artists across nine image-encoder backbones, IDraw consistently outperforms standard image-based verification and reduces verification error by up to 40%. These results demonstrate that inferring drawing behavior from completed images and suppressing drawing content improve digital drawing authorship verification.
SPECTRA: Band-Routed Embedding and Stage-Wise LoRA for Cross-Sensor Fine-Tuning of Geospatial Foundation Models
SPECTRA：用于跨传感器微调地理空间基础模型的带路由嵌入和阶段性LoRA
Li, Xingyan, Caraballo-Vega, Jordan A., Gong, Jie, Carroll, Mark L., Wang, Jianwu
Abstract
Geospatial foundation models (GeoFMs), pretrained on large-scale geospatial data such as Earth observation (EO), climate, and weather data, have shown promising performance when fine-tuned on diverse downstream tasks. However, there are two challenges of adapting EO-pretrained GeoFMs to practical downstream datasets. The first challenge is how to handle spectral mismatch: pretrained patch embeddings expect a fixed set of input bands, whereas downstream sensors may provide different channels. The second challenge is how to reduce fine-tuning cost and make it efficient. While existing work has made efforts on these challenges individually, jointly improving fine-tuning performance under spectral mismatch while reducing adaptation cost remains underexplored. We propose SPECTRA, a parameter-efficient fine-tuning framework that addresses both spectral mismatch and adaptation cost. To handle spectral mismatch, SPECTRA introduces Band-Routed Embedding (BRE), which maps all available downstream bands into the band space expected by the pretrained GeoFM. By using BRE, all available bands in the downstream dataset are utilized to improve the selected-band input without changing the pretrained patch embedding interface. To reduce adaptation cost, SPECTRA further introduces a Stage-wise Transferability-aware LoRA (ST-LoRA) fine-tuning. ST-LoRA estimates stage-wise transferability before fine-tuning and assigns stage-specific LoRA ranks, concentrating trainable parameters on the stages with high transferability for the target task. Across three EO-pretrained GeoFMs and four downstream segmentation datasets, experiments show that BRE improves performance by utilizing all spectral bands, while ST-LoRA reduces trainable parameters compared with full fine-tuning and standard LoRA. Code is available at https://github.com/big-data-lab-umbc/SPECTRA.
Can Urban Blight Be Accessed with Vision-language Models: A Case Study in Detroit
视觉语言模型能否评估城市衰败：底特律案例研究
Yang, Xiaohao, Tian, Aohua, Van Berkel, Derek, Qiang, Xu, Lindquist, Mark
Abstract
Addressing urban blight has seen increased focus in the past 15 years. Assessing urban blight is essential for guiding urban planning, targeting rehabilitation, and safeguarding public health, yet traditional residential blight surveys are difficult to maintain at scale due to the labor-intensive cost and long-term cycle. This study introduced a scalable framework for estimating residential blight using open-source large vision-language models on multiple views. Structured prompts guided models to evaluate housing attributes, including roof integrity, wall damage, and broken or boarded openings, producing both binary assessments and probabilistic estimates of disrepair. To evaluate the performance of these visual assessments, we compared professional human annotations of these features across several models, including an ensemble stacking approach based on XGBoost and a weighted scoring system. Results showed that (i) multiple street views can contribute to the improvement of accuracy, (ii) large vision-language models have different strengths of inference, (iii) the ensemble learner outperforms individual base models, enhancing robustness across all residential conditions and blight assessment. The practical application of the method allows low-cost tracking and management of housing stock conditions, providing a regularly updatable complement to traditional blight surveys.
Adverse weather image restoration aims to recover clear visibility from degraded images in complex weather conditions. Existing works attempt to address this problem by modeling relationships between pixels, however, this paradigm defies the spatially non-uniformity fact of degradations and learns non-discriminative features from semantic-conflict regions. In this paper, we propose SSR, a \textbf{S}emantic-center guilded \textbf{S}tate space model for image \textbf{R}estoration. The key idea of SSR is to shift the conventional scanning strategy of pixel-serial to semantic-guilded one. Specifically, we introduce a Superpixel-guided Selective Scan Mechanism ($\text{S}^3$M), which first partitions the image into perceptually coherent regions via superpixel clustering and then performs relations modeling within the semantic-related regions. Moreover, a Region-level Gating Mechanism (RGM) is developed to perform intra-region calibration by modulating degradation outliers within each semantic superpixel unit along the channel dimension. Extensive experiments on \textbf{6} well-established benchmarks demonstrate that SSR performs favorably against state-of-the-art models with competitive computational cost.
End-to-end (E2E) autonomous driving algorithms require rigorous closed-loop validation in simulation environments offering high visual fidelity, strong interactivity, and real-time performance. Existing approaches, from game engines to static neural rendering, inherently trade off these requirements and struggle with the dynamic scene composition essential for E2E testing. To bridge this gap, we propose a novel decoupled 3D Gaussian Splatting (3DGS) framework tailored for large-scale E2E evaluation. We fundamentally decompose scenes into a high-fidelity static background and manipulable dynamic agents using an object-centric canonical representation. To resolve resulting representational conflicts, we introduce three targeted modules: (1) asset compression via perceptual pruning and vector quantization for real-time traffic rendering; (2) map-guided geometric registration leveraging semantic topology to strictly align trajectories; and (3) proxy-based relighting transferring ambient illumination for seamless photometric integration. Extensive experiments demonstrate that DecoupleGS achieves a balanced fidelity-efficiency trade-off, improves metric and photometric consistency, and provides a practical closed-loop sensor simulation platform for E2E autonomous driving evaluation.
We present LiveLight, the first diffusion-based framework for real-time streaming video relighting with interactive 3D lighting control. Achieving this is non-trivial, as it requires overcoming three critical challenges: effectively injecting dynamic 3D lighting into a diffusion model, maintaining high-fidelity generation under an extremely low NFE (Number of Function Evaluations) budget for real-time speed, and facilitating continuous streaming for interactive control. To address these pain points, we propose three key designs. First, for accurate lighting injection, we propose a lightweight adapter that feeds Multi-Plane Light Irradiance (MPLI) conditions-depth-aware irradiance maps encoding 3D lighting geometry-directly into the diffusion backbone. Second, to prevent rendering quality degradation at low NFEs towards real-time distillation, we introduce a geometry-guided feedback branch. This training-time constraint leverages a frozen geometry estimator to enforce depth- and normal-consistent relighting, ensuring geometrically plausible shading without adding inference overhead. Finally, to enable streaming interaction, we develop a progressive rolling-window strategy that maintains a denoising ladder of latent chunks at varying noise levels. By propagating intermediate states, this strategy guarantees temporal coherence and supports arbitrarily long video relighting with per-frame reference refresh. Extensive experiments on real-world and synthetic benchmarks demonstrate that LiveLight achieves state-of-the-art relighting quality while running at real-time speed, significantly outperforming offline baselines in temporal stability, lighting controllability, and user preference. To foster real-time interactive relighting research, we will publicly release our models, training data, and synthetic data generator.
Investigating Social Bias in Narrative Image Generation
叙事图像生成中的社会偏见研究
Park, Junyeong, Min, Sowon, Jang, Euna, Kim, Soobin, Jin, Jiho, Lim, Hyunseung, Bae, Gahyeon, Hong, Hwajung
Abstract
Text-to-image (T2I) generation models are increasingly embedded in applications such as media content creation and education, raising concerns about how their outputs may reproduce social biases. Prior work has shown that T2I models exhibit social biases, yet existing evaluations largely focus on a photo generation task. As a result, it remains unclear whether and how such biases manifest in more narrative visual formats, such as storyboards and comics, where characters and events are presented across multiple panels. In this work, we compare bias expression across photo, storyboard, and comic generation in six T2I models by adapting BBG, a text-based bias evaluation framework, to image generation. Our results show that proprietary models generate 25.9% biased outputs in photo generation on average, with biased outputs increasing by 9.6pp in storyboard generation and 18.2pp in comic generation. We also find that photos mainly encode biases through subtle visual cues, while storyboards and comics reveal them more explicitly through event sequencing, character positioning, narrative resolution, and textual elements. These findings show that biases that remain less visible in photo generation may surface in narrative visual formats, highlighting the importance of evaluating T2I systems with diverse visual formats beyond photo generation.
Universal Multimodal Embeddings (UMEs) aim to unify various modalities and tasks into a shared representation space. In recent years, this field has witnessed substantial progress driven by the development of Multimodal Large Language Models (MLLMs). However, a crucial capability, visual identity discrimination, remains underexplored in existing UME methods, despite its critical role in a wide range of tasks, including instance retrieval, re-identification, and identity preservation in AI-generated content. To bridge this gap, we propose a unified formulation for visual identity discrimination~(VisID) and introduce $\textbf{MVEB}$ ($\textbf{M}$ultimodal $\textbf{V}$isual Identity $\textbf{E}$mbedding $\textbf{B}$enchmark), a large-scale benchmark curated from both real-world and synthetic datasets to support evaluation and training. Furthermore, we present a simple yet effective learning framework that jointly optimizes general multimodal and visual identity representations through a carefully designed identity-aware sampling mechanism. Extensive experiments demonstrate that our approach successfully endows UMEs with strong identity discrimination capability and maintains competitive general multimodal performance. We believe this work not only illuminates a critical yet neglected capability, but also takes a step toward more holistic universal multimodal embeddings. Code and data are available at \href{https://chrisclear3.github.io/MVEB}{MVEB}.
Existing RGBT trackers typically employ fusion functions with fixed parameters across different targets and scenarios. Although dynamic-architecture methods improve fusion flexibility by selecting among predefined operations, they still cannot adapt the fusion parameters to the evolving target state. To address these issues, we propose a Parameter-Dynamic Adaptive Fusion and Calibration Network (PAFCNet) for RGBT tracking. PAFCNet dynamically generates target-conditioned parameters for multimodal fusion and temporal calibration, enabling the tracking process to adapt to target appearance variations and modality quality fluctuations. Specifically, we introduce a Target-Adaptive Hypernetwork (TA-HyperNet) that leverages template representations, which preserve stable target identity and recent appearance changes with less background interference, to generate target-conditioned parameters for subsequent fusion and calibration. Based on TA-HyperNet, we design a target-aware parameter-dynamic fusion module that uses the generated parameters to modulate the fusion process. This enables the fusion module to adapt to changes in target appearance and complex scene conditions. Furthermore, since spatio-temporal information propagation may accumulate tracking noise, we propose a dynamic spatio-temporal calibration module that employs TA-HyperNet to generate calibration parameters for spatio-temporal tokens. By dynamically calibrating historical information before propagation, the module improves the reliability of temporal representations. Experimental results demonstrate that PAFCNet achieves competitive performance on multiple RGBT tracking benchmarks.
Deep learning models for medical image analysis typically apply a fixed amount of computation to every input, regardless of case difficulty. Anatomy-guided dual-stream architectures have been shown to improve diagnostic performance, but they evaluate both streams unconditionally, even on cases a single stream could already resolve confidently. We propose SecondOpinion, a framework in which a fast primary stream processes every case, while a second, anatomy-guided stream is invoked only when GateKeeper, a gating mechanism trained explicitly as a binary correctness classifier, judges that the primary stream's prediction needs additional scrutiny, much as a clinician might seek a second opinion on a difficult case. When activated, the two streams are combined through a lightweight cross-attention fusion module. We evaluate SecondOpinion on a unified five-class chest X-ray dataset and a pelvic fracture dataset, the latter including a held-out, harder subset of fractures that are invisible on X-ray but confirmed via CT. SecondOpinion matches or exceeds prior state-of-the-art performance on both tasks, while activating its anatomy-guided stream on only 9.23% of chest X-ray cases, rising to 24.12% on visible fractures and 45.71% on invisible fractures, an activation rate that tracks task difficulty directly. These results suggest that supervising a gating signal toward correctness, rather than relying on unsupervised confidence, allows a model to allocate anatomical reasoning where it is actually needed.
Diffusion vision-language models (dVLMs) iteratively denoise masked responses while conditioning each denoising step on visual evidence, making visual conditioning a substantial recurring inference cost. Unlike autoregressive decoding, diffusion generation repeatedly revisits the entire response as uncertainty evolves. Our analysis reveals that visual evidence demand is strongly step-dependent, motivating adaptive allocation across denoising steps. Existing inference acceleration methods operate through decoding-side strategies or visual token compression via pruning and merging, but do not explicitly treat visual evidence as a resource whose demand evolves across the diffusion process. Therefore, we present Denoising-Aware Visual Evidence Trajectory Allocation (DAVET), a training-free framework that allocates visual evidence according to the evolving generation state. Starting from a phase-conditioned evidence trajectory, the proposed allocation policy uses operation demand to set an evidence reserve whose allocation at each denoising step is modulated by trajectory risk. DAVET realizes the resulting budgets through a hierarchy of evidence views constructed from a single visual encoding, separating when and how much evidence is needed from how the evidence views are constructed. Evaluated on two representative dVLMs, LLaDA-V and LaViDa, across multiple visual-understanding benchmarks, DAVET achieves an average speedup of 1.55$\times$ with an average relative performance drop of 1.86\%, showing that denoising-aware visual evidence allocation can reduce visual conditioning cost while largely preserving generation quality.
Detail Continuation over a Trustworthy Coarse Scale for Autoregressive Super-Resolution
基于可信粗尺度的自回归超分辨率细节延续
Fang, Hongyi, Wu, Jiahui, Yue, Yichen, Zhou, Benjia, Zeng, Dan
Abstract
Hallucination remains a persistent challenge in generative super-resolution (GSR), where reconstructed results may contain visually plausible yet weakly supported content, structural deviations, or unnatural textures with respect to the low-resolution (LR) input. Existing GSR methods have extensively explored the trade-off between perceptual realism and reconstruction fidelity, but the division between preserving reliable coarse-scale information and restoring more uncertain fine details is often handled implicitly within the overall restoration process. Visual autoregressive (VAR) modeling provides a natural opportunity to revisit this issue, as its coarse-to-fine next-scale prediction offers an explicit scale-wise generation interface. However, existing VAR-based SR methods still inherit the original full 1-to-$N$ autoregressive generation path, even though, for super-resolution, coarse-scale information in LR is often relatively more reliable, while long autoregressive chains may accumulate prediction errors. Motivated by these observations, we propose \textbf{K2N}, which reformulates VAR-based SR from full-path generation into a $k$-to-$N$ detail continuation process. Specifically, early coarse-scale states are established directly from LR, while only the remaining finer scales are restored autoregressively. Experimental results show that K2N remains competitive with the VARSR baseline on standard SR metrics, while exhibiting clearer advantages on hallucination-focused evaluation. These findings suggest that explicitly rethinking the generation path in a scale-wise manner can be a promising direction for improving the reliability of generative super-resolution. Our code will be released soon at https://github.com/BRL-SYSU/K2NSR.
Part-level 3D generation has recently attracted increasing attention for producing structured and editable 3D assets. However, existing methods typically decompose objects according to functional semantics rather than the editable material boundaries (e.g., fabric, wood, metal) required in practical 3D applications such as interior design. Additionally, current methods often generate parts independently, causing computational costs to scale linearly with the part count. To address these limitations, we present PartMat, an efficient material-aware 3D part decomposition pipeline that represents multi-part geometry with a single global latent. Given a reference image and a single whole-object geometry, PartMat decomposes the object into parts that follow material boundaries. First, we propose PartVAE to learn such a unified representation and decode all material parts in a single forward pass, thereby decoupling inference cost from the number of parts. Second, with this representation, a diffusion model is trained for part generation and refined via reinforcement learning for accurate material assignment and overlap suppression. Finally, to recover fine-grained geometric details, we introduce a sparse-voxel flow-matching model with part attention for geometry post-processing. Extensive experiments demonstrate that PartMat significantly outperforms existing baselines in material-aware decomposition accuracy and achieves comparable geometric quality, while maintaining efficient inference.
Multimodal large language models (MLLMs) have advanced visual understanding and reasoning, yet their static parametric knowledge limits their ability to address knowledge-intensive and dynamically evolving open-world problems. To move beyond this limitation, multimodal deep search has emerged as a key direction for open-world information access, evolving from single-turn factual retrieval toward long-horizon, multi-turn search guided by visual evidence. However, existing methods typically confine vision to the input or answer stage, overlooking its role in intermediate reasoning, and lack designs tailored to long-horizon interaction. Consequently, visual evidence rarely drives continued retrieval, constraining both interaction depth and reasoning span. To address these limitations, we propose DeepVoyager-VL, a long-horizon multimodal deep-search framework for vision-in-the-loop search. Specifically, we construct a multimodal event graph to drive data synthesis, yielding problems with intermediate visual dependencies and long reasoning chains. We then design an agent framework for active visual acquisition and on-demand image loading. Finally, we fine-tune models on the synthesized data without reinforcement learning. Extensive experiments across ten multimodal search benchmarks demonstrate the effectiveness of our method.
Real-world video arrives hazy, rainy, dark, or noisy, and a deployable restorer faces three demands at once: no degradation label, native 4K output, and stability in playback. Existing methods answer them separately and break on the joint problem, because per-frame degradation readings flip between frames, downsampled proxies erase the rain and noise they are meant to remove, and dense temporal alignment does not fit 4K memory. No paired benchmark even poses that problem, so we build one. UHV-4K-AIO renders physically modeled haze, rain, sensor noise, and low light over the same 100 clean 4K clips with shared depth and motion, and its construction exposes the split MoCRA is built on: haze and low light survive aggressive downsampling, while rain and noise exist only at native scale. Band-matched compositional conditioning follows, spending conditioning capacity, computation, and supervision in the band where each degradation lives. One dictionary of rank-1 atoms, recomposed sparsely per frame, conditions both a once-per-clip coarse branch and a shallow native-resolution refiner, in 3.6M parameters and with no optical flow. Trained once for all four tasks, MoCRA takes the best task-mean PSNR of eleven retrained image and video baselines, holds warping error at the level of the flow-based video models while never estimating motion, and restores native 4K in under half a second, against 1.7 seconds for the fastest baseline.
Recent visual-text compression (VTC) methods, typified by DeepSeek-OCR, report impressive high token compression ratios for long-context modeling tasks by leveraging text-to-image rendering. However, existing evaluation protocols heavily rely on downstream task performance. Such evaluation metrics fail to accurately measure text preservation due to the strong inherent linguistic priors of Multimodal Large Language Models (MLLMs). In this work, we introduce a new evaluation framework that decouples MLLMs' capabilities to faithfully assess VTC quality. Within this framework, we further introduce the ZeroSense Benchmark to ensure low semantic correlation of testing samples. By eliminating textual dependencies, our benchmark guarantees that the evaluation results are purely reflective of VTC quality, unaffected by the semantic inference capabilities of downstream models. Extensive experiments across multiple datasets demonstrate that VTC quality and downstream task accuracy diverge significantly, highlighting the necessity of our decoupled evaluation framework.
With the widespread adoption of Vision Transformers in modern AI, the need to analyze their inherent representational behavior has become increasingly important. While most existing studies emphasize token geometries and training dynamics, the evolution of representational covariance structures and class-level geometric organization remains comparatively underexplored. In this work, we investigate semantic geometry and class separability as representations evolve across the layers of ViT-Small/16 through TGO-III: Semantic Geometry Observatory. It is a framework designed to analyze the emergence of semantic organization, feature evolution, and class-wise representation geometry throughout training. The framework employs multiple complementary observatories, including Linear Probe Accuracy, Fisher Ratio, Class Centroid Distances, Local Intrinsic Dimension, and Local PCA Rank, to quantify the progressive evolution of discriminative representations. Our analysis reveals that class representations become progressively more linearly separable, Fisher discriminability increases, class centroids move farther apart, and local representation manifolds exhibit structured class-dependent geometric complexity. These observations provide empirical evidence supporting the Semantic Expansion Hypothesis, suggesting that the manifold expansion observed in previous observatories is accompanied by the progressive organization of representations into increasingly discriminative semantic structures. Collectively, TGO-III extends the Transformer Geometry Observatory framework by establishing a direct connection between manifold geometry, covariance evolution, and semantic organization during Transformer training.
Low-light image enhancement (LLIE) seeks to restore structural fidelity, natural color rendition, and proper exposure from images captured under inadequate lighting conditions. Recent state-of-the-art approaches, such as CIDNet, adopt a dual-branch architecture comprising a chrominance (HV) branch and an intensity (I) branch to separately model decoupled chromatic and luminance information within the HVI color space. However, these methods overlook the mutual interaction between intensity and chrominance components, which inherently limits their representational capacity and leads to suboptimal enhancement performance. To address this limitation, we propose the Conditional Mutual Information-Guided Network (CMIG-Net), which leverages conditional mutual information as a principled metric to quantitatively assess the contribution of chrominance features conditioned on the available intensity information. In particular, we design a Conditional Mutual Information Calibration (CMIC) module that generates a conditional information map, enabling region-adaptive recalibration of chrominance representations according to local illumination statistics. Furthermore, we introduce a Dynamic Dual-branch Information Restoration (D2IR) module, which adaptively governs bidirectional information flow between the intensity and chrominance branches, guided by both the conditional prior and the instantaneous restoration state. Extensive experiments on paired LLIE benchmarks demonstrate that CMIG-Net consistently outperforms CIDNet, achieving up to a 0.619 dB gain in PSNR, with a 0.382 dB improvement specifically on the challenging Sony-Total-Dark dataset.
GeoCore-9B: Towards Geo-Aware Generative Foundation Models in Earth Observation
GeoCore-9B：迈向地理感知的地球观测生成基础模型
Do, Jeonghyeok, Kim, Munchurl
Abstract
Existing generative models for earth observation (EO) predominantly rely on fine-tuning natural image priors, which limits their scalability and introduces perspective biases that conflict with geospatial constraints. To address this, we introduce GeoCore-9B, a 9-billion-parameter generative foundation model, which is the first of its scale to be trained from scratch exclusively on EO data. Unlike previous EO foundation models, GeoCore-9B is built upon a Flow Matching-based Diffusion Transformer (DiT) and natively conditions generation on text descriptions and continuous geospatial metadata, including ground sample distances, latitudes, and longitudes. To overcome the convergence and spatial disorientation challenges of training at this scale, we propose a Geospatial Semantic Alignment loss. This objective distills structural Earth surface priors (e.g., terrain and urban areas) from a frozen specialist teacher network, constraining the diffusion latent trajectory during training without adding inference overhead. Pre-trained on the global-scale Git-10M dataset, GeoCore-9B demonstrates strong downstream versatility. Beyond standard proxy generative tasks, we show that GeoCore-9B can be effectively adapted for practical EO applications, including highly challenging tasks such as cloud removal and SAR-to-optical cross-modal translation. Extensive evaluations confirm that GeoCore-9B establishes new state-of-the-art performance in both visual fidelity and geographic structural accuracy.
SpatioLM: Towards General Physical Spatial Intelligence in Vision-Language Models
SpatioLM：朝着视觉-语言模型中的通用物理空间智能迈进
Wu, Jing, Wu, Jianhua, Guan, Jiayi, Chen, Jiahong, Lu, Jinghui, Ye, Hangjun, Gao, Bingzhao, Chen, Long
Abstract
Vision-Language Models (VLMs) perform well on commonsense reasoning tasks but struggle with visual spatial reasoning. Most existing solutions introduce extra 3D prior inputs or external spatial encoders, which increase complexity and degrade the underlying VLMs' general-purpose capabilities after spatial fine-tuning. To this end, we propose a parameter-efficient \textit{\textbf{Spatio}-vision \textbf{L}anguage \textbf{M}odels (SpatioLM)}, that enhances spatial intelligence without extra 3D prior inputs or third-party spatial encoders. Concretely, we design a plug-and-play and non-invasive spatio-vision module that elicits the spatial knowledge inherent in VLMs. Furthermore, we innovatively leverage pseudo depth and camera information as supervision to guide the model in learning physically coherent representations. Extensive experiments show that SpatioLM achieves significant improvements in diverse tasks, including spatial perception and understanding while effectively limiting the degradation of general capabilities. Notably, the model achieves an impressive score of 71.6 on the VSI-Bench (the first model to surpass 70). In addition, it attains competitive performance when transferred to embodied manipulation tasks. Code is available at \href{https://github.com/xiaomi-research/spatio-lm}{\faGithub~spatio-lm}.
Hand-object interaction (HOI) is a fundamental human behavior with broad applications in AR/VR, digital humans, and embodied interaction. Existing methods typically require predefined object geometry, object trajectories, or task-specific conditions, limiting their use with natural real-world inputs. To address this, we study a more practical problem of synthesizing 3D hand-object interaction sequences from a single RGB photograph and an open-vocabulary language instruction, and introduce PhotoHOI. PhotoHOI first uses a vision-language model to parse the input image and instruction into a structured task specification, including the interaction object, target region, and spatial relation. It then recovers a compact task-relevant 3D scene and plans a smooth collision-aware object trajectory based on the recovered object states, support relations, and surrounding scene geometry. To synthesize hand motion that generalizes to real-world photographs and unseen objects, it learns transferable task-conditioned contact and contact-conditioned grasp priors from large-scale affordance and HOI data. The grasp is further refined in a learned latent space, constraining the optimization to a plausible hand-pose manifold. Experiments on GRAB and H2O demonstrate improved contact quality and reduced penetration over representative baselines. Results on real-world photographs further demonstrate higher task success and scene consistency, together with generalization to unseen objects and open-vocabulary instructions.
Rapid and accurate post-disaster building damage assessment is essential, yet remains a challenging task. Unmanned Aerial Vehicle (UAV) imagery offers a timely and high-resolution view of affected areas, but existing Computer Vision (CV) models often demand large annotated datasets, generalize poorly across geographic regions and their assessment policies, and are confined to the specific tasks they were trained for. Large Vision-Language Models (LVLMs) offer a promising alternative through their strong reasoning and generalization capabilities, but fall short on precise, low-level perception tasks such as object detection and accurate bounding box generation. Furthermore, they often require a substantial amount of data for effective fine-tuning on domain-specific tasks. In this paper, we propose a hybrid framework that decouples detection from damage assessment, combining the precision of CV models with the reasoning power of LVLMs. A CV model first detects buildings and generates bounding boxes on the image that are then passed to an LVLM for damage classification and contextual interpretation. We evaluated our framework on two real-world benchmarks: RescueNet and FloodNet. In particular, the best combination under this framework accurately counts intact, partially damaged and completely destroyed buildings, surpassing isolated baselines by up to 2.1 R^2 points, while requiring only limited annotated data for the detection stage. Beyond reporting aggregate gains, we provide a detailed analysis of failure scenarios and edge cases, offering practical insights for practitioners and concrete directions for future work. Our source code and data are publicly available to the research community via the following repository: https://github.com/ungquanghuy-kddi/VLM_GDINO.git
PNEC-Mamba: Prototype-Guided Positive-Negative Evidence Calibration for Hyperspectral Image Classification
PNEC-Mamba：基于原型引导的正负证据校准用于高光谱图像分类
Xu, Mingzhen, Xu, Can, Wang, Di, Guo, Haonan, Du, Bo
Abstract
In real-world hyperspectral scenes, pixel representations are often ambiguous due to factors such as spectral similarity, mixed pixels, and local context interference, which may simultaneously encode discriminative evidence and interfering information. Existing methods mainly focus on learning more powerful representations or modeling broader contexts, but rarely investigate whether the learned representations provide reliable evidence or introduce interference into classification decisions. To address this issue, we view hyperspectral image classification from the perspective of pixel-level evidence reliability modeling and propose PNEC-Mamba, a prototype-guided positive-negative evidence calibration framework. The framework progressively establishes semantic references, separates class-related evidence from interference, estimates pixel-level reliability, and performs selective calibration. First, a full-image state-space encoder extracts pixel representations, while dynamic class prototypes provide semantic references that evolve jointly with the feature space. Subsequently, positive and negative evidence is derived from pixel-prototype competition, explicitly separating discriminative cues that support classification from confusing signals associated with competing classes. Based on these evidence relationships, a multi-source uncertainty estimation strategy is introduced to assess pixel-level reliability, enabling stronger evidence calibration for uncertain regions. Finally, a full-resolution consistency refinement step is applied to recover local spatial details and improve boundary coherence in the final predictions. Extensive experiments on three benchmark datasets demonstrate that PNEC-Mamba achieves superior classification performance compared with state-of-the-art methods.
Simultaneous localization and mapping (SLAM) based on Neural Radiance Fields (NeRF) enables dense, continuous scene reconstruction. However, existing systems operating with limited online resources struggle to simultaneously construct two types of constraints, namely, compact yet discriminative spatial constraints derived from scene representations and persistent temporal constraints derived from historical observations. To address this challenge, we propose CHOW-SLAM, a dense RGB-D SLAM framework that explicitly constructs these complementary spatial and temporal constraints. Spatially, we propose a compact parametric-hash (P-H) hybrid representation that organizes components based on planes and grids across scales in P and H branches. A unified multi-output decoder further aligns the ray termination distributions induced by TSDF and density, preserving geometry and appearance under a compact parameter budget. Temporally, we propose a complementary overlap-window strategy to prevent optimization from being dominated by short-term overlap or weakly related historical observations. Within a fixed budget, the strategy retains recent frames, selects high-overlap local frames, and introduces temporally distributed historical keyframes. Loss-aware keyframe insertion and bundle adjustment scheduling further adapt optimization to tracking quality. In addition, ORB-based tracking and geometric pose estimation are used for pose initialization, followed by neural rendering optimization to improve tracking stability. Extensive evaluations on multiple datasets demonstrate that CHOW-SLAM outperforms state-of-the-art methods in both scene reconstruction quality and camera tracking accuracy. The source code is available at https://github.com/jinjidexiaohuoban/CHOW-SLAM.
Vision-language models (VLMs) are expected to revise their reasoning when visual evidence changes. Failures to do so are often attributed to insufficient visual attention or contextual inertia, leaving unclear what models reuse instead of recomputing from the current image. We show that evidence-bearing reasoning in a prior chain of thought (CoT) can form a textual shortcut that competes behaviorally with visual recomputation. Across 16 VLMs, a matched counterfactual analysis identifies evidence-bearing content as the most robust carrier of prior-CoT influence. Removing this evidence-bearing content shifts answer preference more than removing length-matched non-evidence context or the final-answer span, with prior control weakening progressively as more stale evidence is removed. Reordering this evidence also weakens prior control, showing that its organization modulates shortcut strength. Beyond the immediate answer, the shortcut can retain residual influence after answer correction: weakening current-image support shifts preference back toward the prior answer, while repeated prior answers and reused premises arise mainly when the shortcut remains active. To limit this influence, we introduce Fresh-State Attention Firewall (FSAF), a training-free intervention that isolates fresh computation from the prior CoT. Across five VLMs, FSAF raises visual update rate from 35.28% to 53.61% and reduces prior-answer rate from 39.22% to 3.67%. Reliable VLM self-reflection therefore requires more than looking again: fresh visual recomputation must be protected from stale textual reuse.
CultureVidBench: Benchmarking Cultural Understanding in Text-to-Video Generation
CultureVidBench：文本到视频生成中的文化理解基准测试
Han, Xianjing, Su, Yuhan, Deng, Yang, Ma, Dong, Tay, Wee Peng, Zhu, Bin
Abstract
Text-to-video (T2V) generation models have advanced rapidly, yet their ability to represent diverse cultural contexts remains underexplored. Existing benchmarks mainly focus on perceptual quality, physical plausibility, and text-video alignment, but do not directly assess whether generated videos capture culturally specific objects, actions, rituals, visible text, or audio cues. We introduce CultureVidBench, a comprehensive benchmark for evaluating cultural understanding in T2V generation. CultureVidBench contains 1,000 curated prompts covering 12 countries, 6 continents, 8 cultural regions, and 14 cultural aspects organized into three categories: material culture, social practice & performance, and ritual & ceremony. Designed specifically for video generation, CultureVidBench emphasizes dynamic and multimodal cultural representation, including social interactions, ritual procedure, and culturally appropriate visible text and audio. We evaluate seven representative T2V models through human user studies and MLLM-based automatic assessment across cultural faithfulness, multimodal cultural rendering, semantic adherence, and perceptual quality. Results show that although current models achieve strong semantic adherence and visual quality, they often fail to faithfully capture fine-grained cultural details, particularly for underrepresented regions, rituals, and multimodal cultural cues.
Controlling human motion and camera movement is essential for faithful human-oriented video generation, yet remains challenging in multi-person scenes with large body motions, occlusions, and dynamic cameras. Existing pipelines typically rely on visual motion sequences, such as skeleton maps, pose maps, or rendered body representations, for motion control, while using camera embeddings for camera control. Such heterogeneous control interfaces force video generation models to reconcile pixel-aligned visual cues with non-visual geometric embeddings, making motion-camera attribution difficult and sensitive to camera estimation errors. We propose \textbf{UniMoCa}, a representation-driven framework that unifies motion and camera controls in visual space. At the core of UniMoCa is \textbf{Motion-Camera Visual Proxy} (\textbf{MCVP}), a mutually-sharable novel representation that converts 3D human motion and camera trajectories extracted from driving videos into an identity-neutral visual proxy. MCVP renders temporally aligned human geometry under the recovered camera trajectory and augments it with explicit camera trajectory markers, replacing heterogeneous visual-parametric controls with distinguishable visual cues. As both control factors are represented in the same visual space, they become mutually compatible rather than heterogeneous, enabling consistent joint reasoning and editing during video generation. We further curate a \textbf{MCVP-Video} dataset covering complex actions, multi-person interactions, and diverse camera trajectories. Experiments based on the Wan2.2 I2V show that UniMoCa achieves substantial gains in human motion control, camera control, temporal consistency, and camera-aware robustness with minimal additional complexity. More details are shown in our Project page: https://tanliming-daniel.github.io/UniMoCa/.
Long-horizon event-based action understanding remains underexplored because existing datasets largely comprise short, trimmed clips, while collecting native event streams with dense temporal annotations is costly. We introduce Event ActivityNet, a large-scale simulated-event benchmark derived from human-annotated, untrimmed ActivityNet videos. It comprises 3,263 videos, 200 action classes, and 106.94 hours, with matched 5-bin and 9-bin event-voxel representations, temporal action annotations, and timestamped captions. The benchmark supports annotated-segment action recognition, auxiliary event-language alignment, and causal online temporal action localization. We generate event voxels directly from non-interpolated source videos in decoded frame order, retain per-video rational nominal or average frame-rate metadata for approximate time mapping, and use action-center reconstruction LPIPS as a soft diagnostic of retained reconstructable content. We establish baselines for adaptive event framing, prompt-caption alignment, and event-only, RGB-only, and RGB-event localization. Under a progressive nested-scale training protocol, recognition Top-1 accuracy increases from 52.25 to 66.42, while online temporal localization average mAP improves from 21.7 to 29.0. Moreover, staged Event ActivityNet pretraining followed by native-event fine-tuning consistently outperforms target-only and joint-from-scratch training across multiple supervision budgets. Event ActivityNet provides a scalable benchmark for long-horizon event modeling, although native-camera evaluation remains essential for deployment-oriented conclusions.
Fixed-layout indoor furniture styling requires selecting assets that form a coherent room without changing the prescribed furniture categories, positions, orientations, or scales. Existing approaches typically retrieve each asset independently or rely on static local relations, making them prone to shape, material, and color conflicts after scene composition. We introduce StyleForge, a scene-level structured selection framework built on a dynamic hypergraph style field. A frozen multimodal large language model extracts structured style priors from an open-ended style request and the fixed layout, while StyleForge maintains a learnable candidate distribution for each furniture slot. Conditioned on the target style, the dynamic hypergraph style field adaptively activates and weights layout-induced hyperedges to capture higher-order dependencies among furniture. Counterfactual style preference learning then treats each candidate as a local substitution in the current style field and evaluates its contextual compatibility using Mahalanobis energies. Training alternates between optimizing the style field and the candidate logits. At inference, the model remains frozen and test-time training updates only room-specific candidate logits, progressively correcting cross-slot style conflicts as the global scene context evolves. Experiments on 3D-FRONT demonstrate state-of-the-art furniture retrieval and scene-level style coherence, producing more coherent fixed-layout furniture arrangements than object- and scene-level retrieval baselines.
4D Gaussian Splatting (4DGS) excels in dynamic 3D reconstruction and real-time novel view synthesis via efficient 4D Gaussian representations and parallelizable rendering. However, existing 4DGS approaches rely on a single polynomial to model motion, which limits performance in complex dynamic scenes where high-frequency motion components are prevalent, and fails to ensure long-term stability due to cumulative trajectory drift. To address these issues, we propose a Fourier Motion Modeling module: this paradigm decomposes motion into frequency-based sinusoidal components, capturing both low-frequency global trajectories and high-frequency local details to model complex motion patterns accurately. It retains the real-time rendering capability of 4DGS while improving complex motion fitting and long-term coherence. Additionally, we integrate a motion-aware regularization strategy into the loss function: it uses frequency-dependent weights to suppress high-frequency jitter while preserving low-frequency motion coherence. Extensive experiments on N3V and Google Immersive datasets from multiple scenarios demonstrate the effectiveness of our method.
Ship detection in Synthetic Aperture Radar (SAR) images plays an important role for maritime situational awareness, especially with respect to different illegal activities at sea such as illegal fishing, smuggling or border violations. Modern ship detection methods using neural networks usually require large training datasets, which are considerably scarcer in the SAR domain than in the electro-optical domain. While several free datasets exist for this task, their availability and usability vary. In this paper, OpenSARShip-Ship Detection Dataset (OSSDD), a new dataset based on the well-known OpenSARShip 1.0 dataset is proposed for training neural networks for SAR ship detection. OSSDD is freely available and contains 15,197 Sentinel-1 amplitude patches in VV and VH polarization, binary ship masks, axis-aligned bounding box and rotated bounding box annotations for a total of 55,759 ships. The construction of the dataset, the contents and structure of the downloadable data and experiments with three common detector models (Faster R-CNN, FCOS, DETR) are shown and discussed. The results serve as benchmarks for future experiments. The dataset is available on Hugging Face at https://huggingface.co/datasets/sylviaHoch/OpenSARShip-Ship-Detection-Dataset.
Chinese Translation
合成孔径雷达（SAR）图像中的船舶检测在海洋态势感知中发挥着重要作用，尤其是在应对海上非法活动（如非法捕捞、走私或边界违规）方面。现代的船舶检测方法通常使用神经网络，且通常需要大量的训练数据集，而在 SAR 领域，这类数据集的稀缺程度远高于电光领域。尽管已有若干免费数据集可用于此任务，但其可用性和实用性各不相同。本文提出了一种基于知名的 OpenSARShip 1.0 数据集的新数据集——开放 SAR 船舶检测数据集（OSSDD），旨在用于训练神经网络进行 SAR 船舶检测。OSSDD 免费提供，包含 15,197 个 Sentinel-1 的幅度图像补丁，涵盖 VV 和 VH 极化，二进制船舶掩膜，轴对齐边界框和旋转边界框注释，总计 55,759 艘船舶。本文展示并讨论了数据集的构建、可下载数据的内容和结构，以及对三种常见检测模型（Faster R-CNN、FCOS、DETR）的实验。结果为未来的实验提供了基准。该数据集可在 Hugging Face 上获取，网址为 https://huggingface.co/datasets/sylviaHoch/OpenSARShip-Ship-Detection-Dataset。
Large language model (LLM) agents increasingly undertake long-horizon tasks that require sustained reasoning, tool use, and revision across many interdependent steps. However, existing agent harnesses maintain task execution, task state, and completion assessment within a growing context, making the state difficult to track and allowing incorrect self-assessments to propagate into later decisions. We reformulate long-horizon execution as a task-state management problem and propose LongHorizon-Harness, which maintains the task state explicitly outside execution and updates it only with facts independently verified from the environment. Its Manage-Execute-Audit(MEA) loop uses a manager to maintain the task state and determine the next subtask, a fresh-context executor to perform it, and a read-only auditor to verify the resulting environment state before the next round. A lightweight AgentAdapter supports interchangeable model and harness backends without modifying their native agent loops. LongHorizon-Harness improves Qwen~3.7-Plus from 51.8% to 80.7% on WeaveBench, from 69.7% to 77.2% on Terminal-Bench~2.1, and from 2.8% to 8.3% on OSWorld~2.0. It also raises Claude Opus~4.7 from 20.0% to 34.3% on an OSWorld2.0 subset, demonstrating consistent gains across models, harnesses, and interaction domains.
SVGEval: A Vision-Grounded Framework for Perceptual-Quality Benchmarking and Evaluation in Text-to-SVG Generation
SVGEval：一个基于视觉的文本到SVG生成中的感知质量基准评估框架
Wang, Yiming, Chen, Ye, Chen, Hanqi, Ni, Bingbing
Abstract
Multimodal large models are increasingly used to generate scalable vector graphics (SVG), but reliable evaluation remains underexplored. Existing protocols are often code-centric or borrow raster-image metrics after rendering SVGs, which fail to reflect human perception and overlook SVG-specific qualities such as geometry and spatial composition. We introduce SVGEval, a vision-grounded multimodal benchmark for human-aligned SVG quality assessment. SVGEval explicitly incorporates visual renderings to evaluate whether models can judge the rendered outcome rather than only inspect SVG code, and provides high-quality annotations obtained via multi-round human labeling with expert refinement. Systematic evaluations across representative multimodal models reveal a clear gap: models perform relatively well on semantic alignment and aesthetics, yet struggle on geometry- and layout-related judgments. Building on SVGEval, we train an explainable SVG quality scorer that outputs multi-aspect scores with textual rationales. Ablations show that explicit visual grounding and reasoning supervision are crucial, especially for spatial and geometric assessment. SVGEval offers a reliable testbed and practical scorer for evaluating and improving SVG generation in the era of multimodal models.
Audio-driven portrait animation has advanced rapidly with diffusion-based generative models, yet real-time one-shot generation with expressive emotion control remains challenging. Existing methods often suffer from insufficient emotion-aware motion priors and expensive appearance computation during multi-step denoising. To address these issues, we propose Proxy Avatar Meets Low-Rank Caching, a cascaded framework for real-time one-shot emotion-controllable portrait animation. Instead of directly generating the target portrait from audio, our method uses a Gaussian-based emotion proxy avatar as a reusable motion generator, which is trained once on a single identity to produce expressive driving videos from audio and emotion labels. Since the proxy avatar only provides motion rather than target appearance or geometry, a large-scale one-shot retargeting model further extracts identity-independent motion from the proxy performance and adapts it to arbitrary target portraits. To improve inference efficiency, we introduce zero-shot appearance reuse with low-rank caching, which caches reference appearance features at the initial denoising step and models subsequent feature variations using lightweight low-rank adapters. Extensive experiments demonstrate that our method achieves stronger emotional expressiveness, better identity-preserving animation, and substantially reduced inference cost, enabling real-time one-shot portrait animation.
Visual token pruning reduces the inference cost of multimodal large language models, but a fixed token ratio is poorly matched to text-rich inputs. In OCR-centric tasks, decisive evidence can be a small number, label, or field whose relevance is specified by the question; indiscriminate pruning can erase that evidence while retaining visually salient but irrelevant regions. We present ET-Prune, a training-free framework that casts pruning as evidence allocation. It derives question-conditioned evidence from a decoder-side partial query-key block, safeguards text-like spatial regions, and converts evidence uncertainty and density into a sample-specific token floor. Three progressive middle-layer events then move the sequence toward this budget, retaining more tokens for diffuse or text-dense evidence and pruning concentrated evidence more aggressively. At the observed point estimates from one deterministic pass per configuration, ET-Prune leads or ties among pruned methods in all six backbone-benchmark comparisons at roughly half tokens. On OCRBench-v2, it leads the strongest pruned baselines by 1.80 and 0.68 percentage points on Qwen3-VL-8B and InternVL3.5-8B, respectively, while retaining about half of the visual tokens; on MMBench v1.1, it reaches 0.8467 circular exact-matching accuracy versus 0.8437 for Vanilla at 54.45% average visual-token retention. These results show a favorable observed quality-cost trade-off for evidence-aware dynamic budgeting in text-rich multimodal inference.
Chain-of-thought (CoT) reasoning can improve performance on difficult video questions but often wastes decoding tokens on simple ones. We study whether a video multimodal large language model can adapt its reasoning effort to each question. We propose AdaThinkV, an adaptive framework for video reasoning that learns whether to reason explicitly without offline difficulty labels, manually tuned confidence thresholds, or an external router. During reinforcement learning, AdaThinkV samples matched rollouts in explicit reasoning and direct answering modes for each prompt. ThinkGain estimates the prompt-level utility of explicit reasoning by balancing its accuracy gain against additional response length, providing supervision for both conditional response generation and autonomous mode selection. For difficult prompts, limited rollout exploration can yield groups in which every response is unsuccessful and accuracy rewards show little variation, providing insufficient signal for learning. We therefore introduce Variance Recovery Policy Optimization (VRPO), which retains and progressively expands these groups to recover informative signals from prompts that are difficult yet solvable. At inference, AdaThinkV selects a response mode and generates the response in a single autoregressive sequence. Across a unified suite of video reasoning evaluations, AdaThinkV achieves a mean accuracy of 40.79 with an average of 257.20 output tokens, outperforming the strongest evaluated adaptive baseline by 2.98 points while using 22.7% fewer tokens. Project page: https://trilarflagz.github.io/AdaThinkV/
DiffPrune: differentiable information throttling for token pruning in vision-language models
DiffPrune：用于视觉语言模型中令牌修剪的可微分信息调节
He, Landi, Yao, Mingde, Young, Shawn, Xu, Lijian
Abstract
Visual token pruning reduces the computational cost of Vision-Language Models (VLMs) by removing redundant visual tokens. The key is to learn a score that measures whether a token is useful. Existing methods typically rely on Gumbel-Softmax to approximate discrete selection during training. Such selectors make the score depend on the behavior of a relaxed pruning operator, not directly on the consequence of information loss. In this paper, we propose DiffPrune, which gives token scores a direct meaning. During training, DiffPrune keeps all tokens and weakens each token's information according to its score. If weakening a token hurts the task, the scorer is pushed to protect it; if not, the token can receive a lower score. Because the loss is differentiated through this actual information-throttling path, the scorer avoids the unstable surrogate path of relaxed token selection. DiffPrune implements this idea with an Information Throttler, which injects variance-preserving noise into visual tokens, where high-score tokens remain close to their original representations, while low-score tokens carry less original information. At inference, the throttler is removed, and hard top-K pruning is applied using the learned scores. Across ten VLM benchmarks, DiffPrune retains 96.5% of full-model accuracy while accelerating LLM prefill by 2.85x, with only 0.69 ms inference overhead. Code will be publicly available.
Rapid advances in image generation models call for interpretable AI-generated image detection methods that not only determine authenticity but also provide supporting visual evidence. Existing approaches may produce inconsistencies between generated explanations and localized evidence regions, undermining the reliability of explanations for authenticity decisions. Meanwhile, existing benchmarks provide limited coverage of the diverse human-centric scenes prevalent in generated imagery. To address these limitations, we investigate authenticity detection with grounded and explainable visual evidence in human-centric scenes. We present HAVE (Human-centric AI-generated Visual Evidence), a diverse human-centric dataset comprising 40K real and 39K AI-generated images from 10 recent generators, with 106K localized evidence instances across 8 evidence categories, each annotated with a bounding box and a region-aligned explanation. We further propose PAVE, a Perception-Aware Visual Evidence framework that jointly performs authenticity prediction, visual evidence grounding, and region-aligned explanation generation. PAVE employs a judge-guided alignment reward to assess region--explanation consistency and evidence validity, together with perception-aware regularization that contrasts token-level predictions between original and randomly masked images to promote reliance on visual input. Experiments on HAVE and external datasets demonstrate strong performance in authenticity detection, visual evidence grounding, and explanation quality. Code and data will be released upon publication.
Denoising diffusion transformers achieve strong generation quality but converge slowly during training. Regularizing their internal representations has emerged as an effective accelerator, yet existing methods split into two families with complementary costs. Target-based methods strengthen representations by aligning them to external features, which requires an external encoder and a learnable projection head to bridge feature spaces. Target-free methods hold no reference at all, and can only repel the model's own features across samples or layers, discarding whatever structure the data contains. Prior work suggests that spatial structure, rather than global semantics, drives the gains of alignment. We therefore ask whether such structure can serve as a target directly, and whether it exists not only within an image but across images. Our key insight is that the clean data latent already carries this structure in the relations among its tokens, where a relation is the similarity between two tokens, a single scalar comparable across feature spaces without a projection head. We propose Structural Parameter-free Affinity Regularization (SPARE), a regularizer that matches the pairwise affinities of intermediate tokens to those of the clean latents. To exploit this structure fully, SPARE extends the matching to token pairs across images, precisely the pairs that prior target-free methods repel by default, and calibrates both relation types with a single learning objective. On ImageNet $256 \times 256$ with SiT backbones under matched 400K-iteration budgets, SPARE adds no encoder, head, or parameters and only 0.08 GB of training memory, yet attains the lowest FID among parameter-free regularizers in every tested setting, recovers 37 to 54\% of REPA's FID reduction, and improves over REPA when combined with it, reaching FID 1.90 under classifier-free guidance at 1M iterations.
Dynamic 3D scene reconstruction has achieved remarkable success under the assumption of strictly synchronized multi-camera inputs. However, in real-world scenarios, temporal asynchrony among capturing devices remains a critical challenge, leading to severe motion blur and geometric artifacts. Existing asynchronous reconstruction methods typically estimate temporal offsets through photometric supervision, but appearance matching provides weak temporal cues under large offsets and complex motions. We attribute this limitation to two major bottlenecks: texture-induced collapse, where low-texture regions provide nearly vanishing alignment signals, and deformation-induced coupling, where temporal errors are absorbed into distorted geometry or motion rather than being explicitly corrected. To address these issues, we propose ASTRA (Asynchronous Spatio-Temporal Reconstruction via Trajectory Alignment), a framework that introduces 2D motion trajectories as explicit, texture-agnostic supervision for asynchronous dynamic reconstruction. Instead of synchronizing cameras solely through rendered color residuals, ASTRA jointly optimizes temporal offsets and dynamic 3D representations by aligning the projected motion of reconstructed 3D points with observed 2D trajectories, while using dynamic and certainty masking to suppress unreliable trajectory constraints. Extensive experiments on different dynamic Gaussian Splatting backbones show that ASTRA preserves high-frequency spatial details and sustains strong robustness even under severe asynchrony with up to 25-frame offsets, achieving approximately 1.4 dB PSNR improvement, reducing temporal-offset MAE by 54.0\%, and nearly quadrupling the synchronization success rate.
Sparse voxel grids preserve the spatial structure needed for detailed 3D reconstruction, but their memory still grows rapidly with resolution as active surface cells increase. We introduce ChunkVAE, a sparse grid variational autoencoder organized around local chunks rather than a global latent volume. Local learned operators permit independently chosen encoder and decoder partitions and allow inference chunk sizes to differ from training. Two complementary data operators make this flexibility practical: Balanced Binary Object Partitioning distributes active cells while limiting replicated overlap, while S-Curve weighted stitching attenuates unreliable boundary features when assembling a global latent or reconstruction. Across three object benchmarks, ChunkVAE is competitive with or better than strong baselines from $512^3$ to $1536^3$; smaller chunks lower peak allocated memory and shorten per-chunk compute, enabling faster parallel inference. Stable stitched latents and improved image to 3D metrics indicate that local compression can scale geometry while retaining the global interface required downstream.
Computer-use agents (CUAs), which empower large language models to autonomously operate operating systems and the web, are increasingly vulnerable to indirect prompt injection attacks. A widely adopted defense is the human-in-the-loop paradigm, in which the agent pauses for explicit user confirmation before executing sensitive operations. While effective against conspicuously high-harm attacks, this defense offers little protection against what we term Invisible Ink Threats: low-harm injected goals, such as starring a repository or installing a package, that are behaviorally indistinguishable from legitimate task execution and thus evade both model safety mechanisms and human oversight. To systematically investigate this blind spot, we present II-Bench, a collection of seemingly harmless adversarial tasks. II-Bench comprises 444 examples targeting confidentiality and integrity attacks across three platforms, spanning three attack categories: page navigation and interaction, sensitive information exfiltration, and code download and execution. Each category is instantiated in both natural language and code forms under two levels of instruction specificity. Furthermore, we construct HITLCUA, a comprehensive adversarial testing framework that integrates a real virtual machine operating system environment with isolated Docker-based web platforms, and simulates human participation by allowing CUAs to consult an API-simulated user before proceeding with suspicious operations. Extensive evaluations of leading CUAs reveal that low-harm injections frequently bypass both agent defenses and simulated user review, exposing severe and previously underexplored security risks in current CUAs.
Remote-sensing videos enable real-time observation of changes in target attributes, short-term activities, and scene evolution. They record motion, actions, interactions, and scene changes that cannot be captured by isolated images. Existing models primarily target single images or discrete temporal observations spanning a long time range. However, a unified evaluation setting for assessing vision-language models on continuous remote-sensing video understanding remains lacking. We introduce RSVideo-10K, a remote-sensing video dataset comprising 10,773 instances, 1.47 million frames, and 17.02 hours of footage, containing both unmanned aerial vehicles and satellite platforms. Its fixed evaluation benchmark, RSVideo-Bench, contains 2,731 test instances and evaluates two complementary aspects of remote-sensing video understanding: L1 Perception and L2 Reasoning, spanning seven capability groups and 17 tasks. Evaluations show that current vision-language models still struggle to recover small local evidence, track short-lived states, and use scene-constrained spatial relations. Based on this analysis, we further propose RSVideo, a reinforcement learning framework for small-target spatiotemporal focusing that selects question-relevant regions across frames and suppresses redundant background tokens. RSVideo achieves a maximum absolute improvement of 9.01% with InternVL3.5-14B and attains the highest accuracy of 40.63% with Qwen3.6-27B across 26 open-source vision-language backbones.Codes will be available at https://github.com/HongjieZhou0329/RSVideo.
Tracking links observations of the same object through visual change, yet cannot by itself determine when the object is empty or filled, intact or cut. We formulate identity-conditioned state-moment retrieval: given a tracked-object history and alternative state descriptions, localize an interval in which each described state holds. Absolute image-text similarity scores descriptions independently; because every visible frame depicts the same target, shared object compatibility can obscure the state evidence needed to identify the target interval. The alternatives provide the missing reference: evidence for one state should be measured against the others. We introduce D\'ej\`a Cue, a training-free framework that turns these alternatives into a vocabulary-relative coordinate system. It subtracts their state-balanced centroid from each description, calibrates frame scores, and scans multiple durations within contiguous visible runs using a frozen encoder. On 78 VOST histories, holding the temporal scan fixed and changing only the query reference nearly doubles R@1 at tIoU 0.5 from 10.3\% to 20.5\% and raises Top-1 tIoU from 16.0\% to 21.5\%. Candidate-rank analyses show that vocabulary-relative queries rank useful intervals higher within the same candidate set. Related state descriptions can therefore serve as an object-specific, query-time coordinate system for reading frozen visual representations.
TBSG-Net: Temporal Bipartite Scene Graph Network for Fine-Grained Video Moment Retrieval
TBSG-Net：用于细粒度视频时刻检索的时间双分图场景网络
Huang, Ji, Dai, Yongsheng, Ren, Tianyu, Devereux, Barry, Wang, Hui
Abstract
Recent advances in proposal-free Video Moment Retrieval (VMR) have highlighted the effectiveness of Static Scene Graphs (SSGs). By modeling objects and their relations at the frame level, SSGs enrich retrieval-oriented video representations. However, integrating SSGs into VMR remains constrained by two inherent limitations: (1) Lack of Temporal Dynamics. SSGs fail to model how objects and their relationships evolve over time, leading to the loss of essential temporal dependencies in video representation; and (2) Lack of Explicit Temporal Span Encoding. SSGs do not explicitly encode the duration of relationships, making precise localization challenging. To address these limitations, we propose Temporal Bipartite Scene Graph Network (TBSG-Net)---to the best of our knowledge, the first Dynamic Scene Graph (DSG) based proposal-free VMR model. Specifically, TBSG-Net leverages DSGs to extract event-centric graph representations of the input video, enabling the modeling of object interactions over time and thus addressing limitation (1). These DSGs are then processed by a novel Dynamic Scene Graph Embedding (DSG-E) module to capture both Temporal Span and spatio-temporal information. First, DSG-E utilizes a TBSG Constructor to transform DSGs into TBSGs, explicitly encoding objects, relationships, and time spans to tackle limitation (2). Second, the resultant TBSGs are passed into a hybrid TBSG Encoder that integrates a Transformer variant for global event modeling and a Graph Convolutional Network for detailed relational reasoning, ultimately producing a more comprehensive spatio-temporal representation. Our experiments demonstrate substantial improvements of TBSG-Net over all baselines.
Recent advances in unified multimodal models have significantly improved text-guided image editing abilities. In particular, models such as Nano-Banana-Pro and GPT-Image-2 demonstrate emerging capabilities in multi-source image editing (MIE), including tasks such as object synthesis, person-background composition, and cross-image style fusion. However, existing benchmarks and image editing assessment (IEQA) methods remain primarily focused on single-image editing tasks and largely overlook the more challenging setting of MIE. This highlights the urgent need for a comprehensive and human-aligned benchmark for MIE. To this end, we introduce MIE-Bench, the first large-scale multiple image editing benchmark with fine-grained human preference annotations. Specifically, MIE-Bench includes 3,000 editing instances across 16 tasks, each involving more than two source images and an editing prompt, together with 36K edited images produced by 12 state-of-the-art editing models and over 108K mean opinion scores (MOSs) covering visual quality, instruction following, and attribute preservation. Based on MIE-Bench, we propose MIEScore, a multimodal large language model (MLLM)-based evaluation model enhanced with skill optimization and multi-dimensional supervised fine-tuning, to provide human-aligned feedback for MIE. Extensive experiments show that MIEScore achieves state-of-the-art performance in aligning with human preferences and generalizes well across other IEQA datasets. Both the dataset and the model are available at https://github.com/IntMeGroup/MIEScore.
Monocular depth foundation models, benefiting from large-scale synthetic training data, have demonstrated strong generalization. However, they often hallucinate depth on non-Lambertian surfaces, estimating reflected content in mirrors or transmitted content behind glass rather than the physical surface itself. Adapting these models with real-world data is challenging because conventional depth sensors are also unreliable in such regions. We observe that while the appearance of a non-Lambertian surface varies with its reflected or transmitted environment, its underlying geometry remains unchanged. Based on this observation, we propose GIFT (Geometry-Invariant Fine-Tuning), a parameter-efficient post-training framework that requires no measured depth labels. We collect groups of RGB images under controlled appearance changes while keeping the camera and target geometry fixed. GIFT exploits geometric invariance across these observations to suppress non-Lambertian depth hallucinations while retaining general depth estimation capability. We further construct a controlled benchmark that evaluates non-Lambertian depth recovery, robustness to appearance changes, and performance retention in other regions. Experiments on our benchmark and an independent real-world dataset demonstrate that GIFT improves depth prediction for mirrors and transparent objects while largely preserving the base model's performance, providing a practical and low-cost approach for adapting monocular depth foundation models to non-Lambertian scenes.
Brain-computer interfaces (BCIs) have been widely used in motor rehabilitation, disease diagnosis, and other neural engineering scenarios. However, conventional neural signal decoding algorithms often suffer from limited generalizability and high adaptation costs, motivating recent interest in BCI foundation models. Existing approaches still struggle to jointly achieve general transferability, accurate decoding, and efficient downstream adaptation. We present STEAM, a hierarchical transfer framework that reconciles general-purpose representation learning with paradigm-specific specialization in EEG foundation models. The framework is instantiated as a dual-branch spatio-temporal encoder in which a shared soft mixture-of-experts (SSMoE) module aligns the spatial and temporal branches, allowing complementary representations to exchange information through a compact set of soft slots. Across seven downstream datasets and fourteen evaluation settings, STEAM attains the best average rank among the compared methods at a competitive inference cost measured in FLOPs. Building upon the Stage-I general initialization, the hierarchical pre-training strategy further specializes the model to a target paradigm without retraining from scratch, yielding consistent gains in paradigm-specific decoding accuracy.
Deep Multimodal Fusion Detection through Spatial Mask and Channel Fusion
通过空间掩模和通道融合的深度多模态融合检测
Wang, Guandi, Li, Ming, Xing, Yunsen, Liu, Junle
Abstract
Deep multimodal fusion for object detection has demonstrated good performance through mining modal characteristics. However, existing feature-level fusion methods mainly weigh between two modalities and unify them in a unified representation space. This can lead to overfitting or over-specialization of the statistical properties of a single modality within a dual-backbone architecture. This paper proposes an Attention-Driven Complementarity Resampling framework for robust improvement of cross-modality object detection. Based on a shared channel spatial attention mechanism, we first introduce the semantic mask exchange to actively mix the boundaries of the modalities during the training phase, forcing the backbone network to learn generalized features without relying on fixed modal labels. Then we propose a learnable channel competition to sample and aggregate features in a channel-wise and learnable way. Our experiments on multiple datasets demonstrate that the proposed method is effective and yields competitive results among existing state-of-the-art approaches. The source code is provided in the supplementary material.
Vision-Text Compression (VTC) renders long texts into images and encodes them through the vision encoder (ViT), compressing thousands of text tokens into far fewer visual tokens. However, since the ViT is pretrained predominantly on natural images, it captures visual attributes (glyphs, font sizes, layout) rather than linguistic semantics, causing rendered-image representations to diverge from native-text representations. We term this cross-path inconsistency and show, via rendering perturbation experiments, that it is a critical yet overlooked bottleneck of VTC. We propose SPIRAL (Self-improving Path Integration and Realignment), a self-supervised alignment framework that closes this gap using only the model's own text-path behavior as supervision, requiring no external teachers or additional annotations. SPIRAL operates at two complementary granularities: token-level on-policy distillation (OPD) for local faithfulness, and sequence-level preference optimization (DPO) for global coherence. On VTCBench, SPIRAL improves the overall score of Qwen3-VL-8B from 35.10 to 54.02, approaching the native text-input performance (55.60) and outperforming models up to 30x larger. The two granularities exhibit complementary strengths: OPD excels at retrieval and is sample-efficient, while DPO is stronger on reasoning and memory and scales better with data. SPIRAL's benefits also generalize to out-of-domain benchmarks, confirming that effective VTC hinges on aligning rendered-image representations back to native-text semantics.
Vision-language models (VLMs) remain unreliable when predictions require fine-grained visual evidence. We identify a previously overlooked cause: spectral response rigidity. Despite substantial frequency variation across images and tasks, pretrained vision encoders exhibit persistent, encoder-specific layerwise spectral profiles that change only marginally under downstream fine-tuning. Since pretrained vision encoders only receive images, they cannot adapt spectral extraction to the evidence required by the current query. We therefore propose HAFI-VLM, which introduces a task-conditioned frequency pathway while preserving the pretrained semantic representation. Hierarchical Adaptive Frequency Injection (HAFI) retrieves complementary low-, mid-, and high-frequency evidence at multiple encoder depths using text-modulated, spatially aligned cross-attention. A Visual Enrichment Layer Adapter further recalibrates shallow LLM attention to effectively utilize the enriched visual tokens. Experiments on LLaVA-1.5 and Qwen2.5-VL demonstrate consistent improvements in general VQA, text-rich understanding, and hallucination robustness, outperforming representation-level enhancement methods and most resolution- or cropping-based approaches without additional high-resolution encoding. Mechanistic analyses show that HAFI restores task-dependent spectral allocation while retaining semantic attention, establishing frequency enrichment as a distinct and effective route for improving VLM perception.
In-context learning (ICL) has attracted increasing attention for enabling models to perform new tasks using only a few ``input--output'' prompt examples. However, existing approaches suffer from \textbf{shallow task adaptation}, where prompts are primarily used as contextual cues to implicitly infer task intent through semantic representations, while the underlying computational process remains unchanged. This limitation restricts task-specific adaptation and compromises inference interpretability. We argue that prompts should not only condition feature representations but also dynamically regulate the model's computation pathways. To this end, we propose \textbf{PromptPath}, an adaptive ICL framework that enables computation-level adaptation through prompt-conditioned dynamic pathways. Specifically, PromptPath introduces a prompt-driven routing mechanism to selectively activate and compose lightweight low-rank experts, forming task-specific computational pathways tailored to different prompts. By integrating prompt information directly into the inference process, PromptPath dynamically reconfigures model computation to enhance task specialization and interpretability. Extensive experiments on 3D point cloud and 2D visual recognition benchmarks demonstrate that PromptPath consistently outperforms state-of-the-art ICL baselines while exhibiting strong cross-domain and cross-task generalization.
Modern vision language models (VLMs) turn high-resolution images into long sequences of visual tokens. Every token traverses the language decoder and persists in its prompt KV cache, inflating inference cost and motivating aggressive visual compression. Existing score-based methods assign each token an independent importance score and retain the Top-K. However, text queries consume collective, signed attention messages from the visual population, not isolated patches. Consequently, equally sized Top-K sets can repeatedly cover one salient region, omit sparse but complementary evidence and discard information carried by the removed population. We therefore formulate faithful visual compression as constructing a compact coreset for decoder messages, and introduce our training-free Grounded Message Coreset Pruning (GMC) which jointly allocates support across query-grounded, appearance, and coordinate-aware evidence, then transports discarded states into selected representatives at their original multimodal positions before physical compaction and native attention resume. This decomposes faithful compression into two coupled components, including selecting carriers that cover the required message modes and realizing the signed population message on those carriers. We further derive bounds connecting their errors to signed-message distortion, visual innovation, and candidate-margin stability. Experiments across multiple VLM families and diverse benchmarks demonstrate strong performance, with GMC-H2 retaining 97.78% Full-relative mean capability on Qwen2.5-VL-7B using 80.2% fewer visual tokens, while GMC-L16 reaches 100.36%. Controlled interventions verify that collective support and population realization jointly drive these gains.
Vision-language models (VLMs) exhibit strong generalization across multimodal tasks but remain vulnerable to adversarial perturbations. Existing attacks typically follow single-trajectory gradient optimization or task-specific objectives, limiting search-space exploration and cross-task transferability. We propose an evolutionary-computation-guided cross-modal attack framework for unified VLMs. The framework adaptively searches both textual and visual spaces. On the textual side, it evolves hard negative semantic embeddings around the source-category representation to provide diverse cross-modal repulsion. On the visual side, it maintains a population of object-region perturbations and combines momentum-based gradient updates with evolutionary selection, mutation, and crossover to more reliably explore multiple feasible trajectories. Jointly optimizing semantic negative guidance and localized perturbations generates adversarial examples that consistently shift source-object semantics toward target categories across vision-language tasks. Theoretical analyses show that the co-evolutionary search preserves perturbation feasibility, prevents degradation of the best observed fitness, and increases the probability of reaching high-margin adversarial regions compared with single-trajectory optimization. Experiments on Florence-2, OFA, and UnifiedIO-2 demonstrate strong overall attack performance across image captioning, object detection, region categorization, and object localization. Ablation studies further verify the complementary effectiveness of text-side semantic evolution and image-side perturbation evolution, as well as the framework's efficiency and cross-task transferability.
HiResNets: Native Full-HD Video Recognition with Foveal Residual Streams
HiResNets：基于注视残差流的原生全高清视频识别
Mall, Shivani, Jain, Swarnim, Henriques, Joao F.
Abstract
Much of the recent progress in image and video recognition has come at the cost of memory: larger models, increased resolution, and longer temporal contexts. An inevitable component is the quadratic (or larger) growth of memory and compute based on image resolution, which is a property of the grid sampling used in convolutional networks and vision transformers. In this work we study residual networks whose convolutional blocks have logarithmic-square growth instead, enabling them to process very high-resolution video quickly. The key insight is to use a residual architecture's residual stream as a high-resolution buffer, to which convolutional blocks only read and write via log-polar image warp operations. Layers adaptively focus on different parts of each frame, with very high resolution only near the focus point. A complete high-resolution representation is built up in the residual stream, analogous to eye saccades creating a complete picture in biological vision, and a theoretical construction is presented that eliminates the quadratic dependency of the residual stream resolution. Experiments demonstrate that our proposed HiResNets learn to foveate around scenes similarly to human vision, and have superior performance in difficult egocentric video recognition tasks, especially egocentric video with small objects and fine-grained recognition.
Division-of-focal-plane (DoFP) color polarization cameras enable snapshot acquisition of color polarization mosaic images, but the inherently sparse sampling pattern makes color polarization demosaicking severely ill-posed. Existing methods often fail to jointly exploit the correlations among polarization channels and the physical constraints inherent in polarization imaging, resulting in noticeable demosaicking artifacts. To address this issue, a quaternion-tensor-based color polarization demosaicking (CPDM) method incorporating Stokes-domain total variation (TV) regularization is proposed. Correlation analysis shows that the correlations among polarization channels are stronger than those among color channels. Accordingly, the color polarization images acquired at $0^\circ$, $45^\circ$, $90^\circ$, and $135^\circ$ are encoded into the four components of a third-order quaternion tensor, with the color channels organized along its third mode. A low-rank prior is then imposed on the quaternion tensor to exploit the global structural redundancy in the color polarization data. Moreover, spatial gradients are mapped to the Stokes domain through an orthogonal transformation to separate intensity, polarization and residual variations, with adaptive quaternion weights enabling component-specific regularization and preserving the energy consistency of the reconstructed Stokes vectors. An efficient optimization algorithm is derived for the resulting model. Extensive experiments demonstrate the superior demosaicking performance of the proposed method.
In this paper, we propose UniqueSplat, a view-conditioned feed-forward 3D Gaussian Splatting model to reconstruct customized 3D radiance fields for each view query. Existing feed-forward methods such as pixelSplat and MVSplat aim to generate fixed Gaussians across all views of each scene by minimizing the error between rendered views and ground-truth images. However, such fixed Gaussians generally render images from all views and lack the ability to adapt to specific viewpoints, as they do not incorporate target view information when predicting Gaussians. To address this, our UniqueSplat learns the view-conditioned information as a prior and incorporates this knowledge into network parameters, so that Gaussians are dynamically adjusted in accordance with different views. Specifically, we propose a two-branch view-conditioned hyperNetwork to simultaneously learn view-agnostic embeddings and view-specific knowledge, which not only explores the shareable knowledge from various views, but also adapts the model to specific views at test time. Extensive experiments on widely-used datasets including RealEstate10K, ACID and DTU demonstrate the superiority of UniqueSplat over the state-of-the-art methods. Moreover, UniqueSplat encouragingly outperforms existing methods in cross-dataset evaluation, showing its notable generalization ability.
Embodied intelligence and world models require video understanding systems to go beyond recognizing objects and actions and develop an understanding of physical regularities. However, despite their strong performance on general video understanding tasks, current video-language models still struggle to reliably determine whether an observed event conforms to specific physical laws. Existing benchmarks primarily assess the physical quality of generated videos, providing limited support for systematically evaluating and improving the physical-law understanding of Video Large Language Models (VideoLLMs). To address this gap, we introduce PhyCheck, a video question answering dataset organized at two complementary levels of granularity. The coarse-grained subset asks models to determine whether the phenomenon shown in a video conforms to or violates physical laws, while the fine-grained subset further examines whether models can capture physical details responsible for the violation or compliance. We use these subsets as structured supervision to improve physical understanding. In addition, the dataset contains a diagnostic subset with external causal context that reveal hidden factors affecting physical plausibility, assessing whether models can recalibrate their judgments accordingly. Experiments with Fine-tune Qwen2.5-VL show that training with the proposed data substantially improves the understanding of physical-consistency, while evaluations in the diagnostic subset reveal that current models still have difficulty incorporating additional causal conditions into their decisions. These findings highlight the gap between recognizing surface-level inconsistencies and understanding underlying physical mechanisms, and provide a foundation for evaluating and improving physical understanding in Video-LLMs.
Chinese Translation
具身智能和世界模型要求视频理解系统超越对物体和动作的识别，发展对物理规律的理解。然而，尽管当前视频语言模型在一般视频理解任务中表现出色，但仍然难以可靠地判断观察到的事件是否符合特定的物理法则。现有基准主要评估生成视频的物理质量，有限地支持系统性地评估和改善视频大语言模型（Video Large Language Models, VideoLLMs）的物理法则理解。为了解决这一问题，我们引入了PhyCheck，一个在两个互补粒度水平上组织的视频问答数据集。粗粒度子集要求模型判断视频中展示的现象是否符合或违反物理法则，而细粒度子集则进一步考察模型是否能够捕捉导致违反或符合的物理细节。我们利用这些子集作为结构化监督，以改善物理理解。此外，该数据集还包含一个诊断子集，提供外部因果背景，揭示影响物理合理性的隐藏因素，评估模型是否能够相应地重新校准其判断。与Fine-tune Qwen2.5-VL的实验表明，使用所提出的数据进行训练显著提高了物理一致性的理解，而在诊断子集中的评估则显示当前模型在将额外的因果条件纳入决策时仍然存在困难。这些发现突显了识别表面不一致与理解潜在物理机制之间的差距，并为评估和改善视频大语言模型的物理理解提供了基础。
In this paper, we propose a characteristic-aware adaptive network named AdaForensics for deepfake detection. Most existing methods learn a fixed network to detect deepfakes based on carefully-designed network architectures. However, these methods employ the same deepfake detector for all the images despite of various facial characteristic, which fail to provide customized forgery detection for different individuals. To address this, our AdaForensics simultaneously learns characteristic-agnostic and characteristic-specific embeddings, where the detector dynamically adapts to varying faces with our designed hypernetwork on the fly. More specifically, our AdaForensics not only explores the shareable abstractions from various deepfake images, but also adapts the detector to the given characteristic at test time. To achieve this, we propose a two-branch HyperNetwork to learn an adaptive deepfake detector, which automatically adjusts the parameters based on characteristic of the input. Extensive experiments on widely-used datasets including FaceForensics, Celeb-DF and DFDC demonstrate our AdaForensics outperforms the state-of-the-art works.
GSRAIN: Physically Calibrated High-/Low-Frequency Rainfall Synthesis for 3D Gaussian Driving Scenes
GSRAIN：物理校准的高/低频降雨合成用于3D高斯驱动场景
Wang, Fanyu, Zhang, Longgao, Chen, Junyi
Abstract
Existing rainfall simulation methods for autonomous driving remain limited in physical controllability and multi-view consistency. This paper presents GSRAIN, a high-/low-frequency rainfall synthesis method for 3D Gaussian Splatting (3DGS) driving scenes. GSRAIN constructs a high-frequency raindrop model from measured rainfall data and generates low-frequency rainy appearance using a geometry-aware single-step diffusion model. The two effects are then fused in a unified 3DGS scene, enabling rainfall-intensity control over the range of 0--13~mm/h. The proposed method achieves a Fr\'{e}chet Inception Distance (FID) of 149.09, outperforming CycleGAN-Turbo (155.71) and WeatherEdit (157.94). Object-detection and closed-loop driving experiments further show that the generated scenes expose scene-dependent performance changes of the evaluated algorithms under controllable rainfall. These results indicate that GSRAIN provides an effective approach for constructing physically controllable, repeatable, and closed-loop-compatible rainy-weather test scenes for autonomous driving.
SWINSleepNet: A Hierarchical Context-Aware Framework for Sleep Staging (v2)
SWINSleepNet：一种层次化的上下文感知睡眠分期框架 (v2)
Wang, Chongjian, Gao, Junjie
Abstract
Automatic sleep staging is a critical role in sleep disorder diagnosis, sleep quality assessment, and long-term health monitoring; however, existing approaches suffer poor performance on ambiguous and transition-related sleep stages, caused by inadequate modeling of fine-grained intra-epoch structures and complex cross-region spectral dependencies. Traditional epoch-level encoders commonly fail to extract subtle temporal microstructures and intra-epoch cross-region interactions, resulting in unsatisfactory recognition accuracy for hard categories such as the N1 stage. To tackle these drawbacks, we propose SwinSleepNet, a hierarchical context-aware dual-stream framework that separately optimizes intra-epoch representation learning and inter-epoch contextual modeling. Concretely, we characterize each sleep epoch from two complementary perspectives: raw time-domain EEG signal and its time-frequency transformation. The time-domain branch adopts convolutional encoders to capture fine waveform temporal details, and the time-frequency branch uses Swin Transformer to extract local spectro-temporal features, hierarchical multi-scale information and long-range spatial dependencies. The multi-branch extracted features are fused into integrated embeddings, which are optimized by a bidirectional context module to capture cross-epoch temporal dependencies for final sleep stage classification. Comprehensive experiments on Sleep-EDF-20, Sleep-EDF-78 and SHHS datasets verify that our method achieves competitive overall performance, and exhibits stronger robustness and stability on difficult N1 stages and transitional epochs. The results prove that optimized intra-epoch representation learning based on hierarchical architecture greatly benefits automatic sleep staging tasks.
SPIRIT: Spatio-temporal Pairwise Relational Modeling of Instrument-Tissue Interactions for Surgical Action Triplet Recognition
SPIRIT：用于外科操作三元组识别的仪器-组织交互时空成对关系建模
Sharma, Saurav, Arboit, Lorenzo, Banik, Nabani, Meuli, Sarah, Alekseenko, Julia, Liechti, Jan, Heitzinger, Franziska, Orsi, Michela, Mutter, Didier, Gero, Daniel, Nett, Philipp C., Muller, Beat P., Lavanchy, Joel L., Padoy, Nicolas
Abstract
Fine-grained understanding of surgical activity is essential for context-aware assistance in the operating room, including safety monitoring, adverse event identification, and skill assessment. Surgical action triplets, defined as tuples of the form , provide a structured description of instrument-tissue interactions. A key open problem, however, is how to learn triplet representations that remain reliable across institutions, where surgical video varies in acquisition conditions, surgeon style, tool usage, and tissue handling, while existing triplet datasets do not support explicit evaluation of center-wise transfer. To address this problem, we propose \textbf{SPIRIT}, a structured framework for surgical action triplet recognition designed to learn interaction representations that transfer more reliably across centers. Instead of treating each triplet as a flat class label, SPIRIT first learns spatio-temporal representations for instruments, verbs, and targets, then models their pairwise relations, and finally composes them into coherent triplet predictions, with multi-head distillation used to stabilize learning. To evaluate this setting, we establish \textbf{MultiBypass-4C-T40}, a multi-centric dataset for dense surgical action triplet recognition in Roux-en-Y gastric bypass across four geographically distinct centers, with auxiliary phase and step annotations. Across multiple evaluation protocols, SPIRIT consistently outperforms strong recent baselines, highlighting the value of explicit relational reasoning for multi-centric triplet recognition. Code will be available at https://github.com/CAMMA-public/multibypass-4c-t40.
Although image deraining has advanced substantially, existing methods mainly focus on 2D image restoration. As spatial intelligence applications such as embodied AI and autonomous driving continue to emerge, reconstructing clean 3D scenes from sparse rainy views in a feed-forward manner becomes increasingly important. Existing feed-forward 3D Gaussian Splatting (3DGS) methods often assume clean inputs and collapse under rainy conditions. To this end, we present \textbf{\textit{DerainSplat}}, a feed-forward framework that reconstructs clean 3D scenes from only a few rainy views. To support this task, we build a large-scale multi-view derain dataset through a four-stage synthesis pipeline that sequentially models overcast illumination, depth-dependent haze, rain streaks, and lens raindrops, producing privileged weather factors. We introduce a weather net that predicts the weather factors from rainy context and yields two support maps. Scene support modulates cross-view cost-volume matching, while radiance support drives depth-aligned appearance fusion to fill corrupted pixels. The derived geometry evidence further attenuates Gaussian opacity to reduce spurious structures. A rainy cycle consistency re-renders clean views using the predicted factors and aligns them with rainy inputs. Extensive experiments show that \textbf{\textit{DerainSplat}} outperforms existing methods on various datasets, including RealEstate10K, ACID, Mip-NeRF360, and real-world rainy scenes, with strong cross-dataset generalization.
Thermal imaging remains effective under adverse illumination, yet passive long-wave infrared (LWIR) measurements often lack fine texture. Existing thermal texture imaging approaches commonly rely on spectral sensing or registered auxiliary modalities, incurring substantial data throughput or vulnerability to cross-modal degradation. We introduce T$^2$exture, a sparsely perturbed thermal texture imaging framework that aims to reconstruct temporally dense thermal texture sequences from densely sampled passive frames and a few actively perturbed keyframes. We define thermal texture as the residual between a source-on observation and its corresponding source-off passive state. Under sparse LWIR illumination and rapid quasi-steady paired acquisition, this residual attenuates the passive-emission background and approximates a source-induced reflected response, exposing localized material- and geometry-dependent texture. T$^2$exture reconstructs a dense sequence of this source-conditioned response through two stages. Stage 1 estimates the unobserved source-off passive state at each active instant from neighboring passive frames to obtain reliable differential texture anchors. Stage 2 combines sparse anchors with passive structural context near each target time to reconstruct the dense sequence. On the simulated benchmark, T$^2$exture adds only 0.20M parameters to AMT-L while improving PSNR by 6.66 dB. Extensive evaluations on simulated and real acquisitions further show clearer texture recovery and stronger structural preservation than representative VFI baselines. These results establish T$^2$exture as a practical framework for thermal texture imaging under sparse active acquisition.
RSC-GestureNet: Reliability-Aware Selective Causal Recognition of Chinese Traffic Police Gestures
RSC-GestureNet：可靠性感知的中国交通警察手势选择性因果识别
Li, Cheng, Gao, Renjun, Fu, Boyi
Abstract
Traffic police gestures are safety-critical perception cues for autonomous driving. A deployable recognizer must infer commands causally from continuous full-frame video, remain stable around transitional arm motion, and avoid over-trusting corrupted pose measurements. This study presents RSC-GestureNet, a reliability-aware selective causal recognizer, for Chinese traffic police gestures. The model treats pose confidence as a first-class signal: unreliable joints are down weighted during graph reasoning, temporal evidence is aggregated causally, and calibrated predictions are selectively emitted through a reliability-aware inference rule. We further introduce CTPGesture-C, a reproducible feature-level corruption benchmark with seven pose/RGB degradation families, and an RGB-level diagnostic in which corrupted frames are reprocessed by MediaPipe before recognition. On the complete official CTPGesture v1 split (134,424 labeled frames and 33,451 causal windows), RSC-GestureNet achieves 93.33+-0.24% accuracy, 91.71+-0.27% macro-F1, 91.69+-0.29% online macro-F1, 98.80+-0.07% Early@10, 0.153+-0.013 s TTC, and the best robust macro-F1 among evaluated methods. Under the same split and causal protocol, it exceeds reproduced traffic-specific MD-GCN and HLP-GCN baselines by 3.23-4.11 macro-F1 points and 2.15-3.07 online-F1 points. These results, together with calibration, selective-risk, statistical, adaptive-branching, and image-level re-extraction analyses, indicate that explicit pose-reliability modeling improves early, stable, and robust traffic-command recognition.
Sparse-view 3D Gaussian Splatting Super-resolution is highly challenging since the sparse and low-resolution (LR) inputs lack sufficient geometric and high-frequency information for accurate reconstruction. To achieve high-quality reconstruction, existing sparse-view super-resolution methods adhere to two-stage pipeline that performs LR Gaussian reconstruction and then high-resolution (HR) Gaussian refinement, which directly results in stage-wise Gaussian transfer and reconstruction error accumulation. To this end, we propose CLEAR, a Conflict-aware Learning via Evidence-guided Adaptive Routing, as the first unified single-stage framework for Sparse-view 3D Gaussian Splatting Super-resolution. Specifically, CLEAR performs joint the optimization of authentic LR observations and external HR priors within a unified Gaussian representation. To mitigate the gradient conflicts introduced by sparse supervision during training, we propose a Gaussian-wise conflict-aware optimization strategy that regards the LR gradient as a reliable anchor and applies evidence-conditioned soft correction only to severe HR conflicts. Moreover, to recover high-frequency details, we introduce an evidence-guided Patch-to-Gaussian routing mechanism which estimates patch reliability and detail demand, lifts them into Gaussian space, and selectively routes high-frequency gradients and densification. Finally, we employ shared Gaussian dropout and a detached mid-training anchoring to enhance the robustness of training framework. Extensive experiments on both synthetic and real-world $4\times$ super-resolution benchmarks demonstrate that CLEAR consistently achieves state-of-the-art rendering quality and superior geometric fidelity.
Whole-body dual-energy X-ray absorptiometry (DXA) scans are routinely acquired to measure bone density and regional body composition, leaving their spatial structure largely unused. Here, we show that self-supervised learning (SSL) can convert raw DXA images into representations of systemic health. We introduce LeDXA, a vision model based on a joint-embedding predictive architecture (JEPA) that learns by predicting latent representations rather than reconstructing pixels. Trained from scratch on 11,540 unlabeled Human Phenotype Project scans, LeDXA was evaluated internally and on 47,400 external UK Biobank (UKBB) scans. It improved cross-cohort prediction of prevalent diseases and biomarkers beyond scanner-derived DXA measurements and DINOv3, a state-of-the-art general-purpose model, despite approximately 150,000-fold fewer training images and nearly 40-fold fewer parameters. Over a median 4.3-year UKBB follow-up, LeDXA improved incident disease prediction over tabular DXA measures, with the largest gains for hip and knee arthrosis and type 2 diabetes. For hip arthrosis, 66% of incident cases occurred in the highest-risk quartile versus 41% for tabular measures. Its representations predicted chronological age externally (r = 0.88; mean absolute error = 2.90 years), and the biological-age gap tracked broader disease burden and a 45% higher mortality hazard in the oldest-appearing quartile. The gap also decreased in women after starting hormone-replacement therapy, suggesting it may be modifiable. Genome-wide associations recovered mostly known body-composition and bone-density loci, and LeDXA embeddings were more heritable than DINOv3's. These findings reveal prognostic information in DXA images that conventional readouts discard, learnable with relatively little data and modest compute.
VARPose: Flexible 2D Pose Densification via Visual Autoregressive Modeling for Enhanced 3D Lifting
VARPose：通过视觉自回归建模实现灵活的2D姿态稠密化以增强3D提升
Pu, Kaiyuan, Yang, Tiantian, Zeng, Dan
Abstract
Visual AutoRegressive Modeling (VAR) has excelled in natural image generation via next-scale prediction, but its use on topology-structured data like human skeletons is still unexplored. VARPose is proposed to adaptively densify 2D sparse poses, thereby enriching the anatomical information available for 3D lifting models. Our core contributions are twofold. First, we introduce a Granularity-agnostic Pose Tokenizer (GPT), which employs a single hybrid codebook and a residual quantization strategy to encode poses of varying densities into a unified, multi-scale discrete representation. Our results demonstrate the strong generalizability of this representation. By decoupling the representation from the projection, we can successfully decode novel pose granularities using a frozen codebook with a retrained decoder. Second, we propose UniSkelar, a unified autoregressive model that treats "joint density" as "scale". UniSkelar learns to predict the token sequence for the next density level in a coarse-to-fine manner, conditioned on the sparsest pose. VARPose not only outperforms state-of-the-art methods and generalizes to unseen granularities, but also confers tangible performance gains on downstream tasks, such as 3D Pose Estimation and Human Mesh Recovery, through 2D pose densification. Our code and model are available at https://github.com/BRL-SYSU/VARPose.git.
Vision-language models (VLMs) such as CLIP exhibit remarkable zero-shot capabilities, yet their performance frequently degrades sharply under unexpected test-time distribution shifts. While Test-Time Adaptation (TTA) offers a promising solution, continuously adapting VLMs over an unlabeled test stream presents fundamental challenges. Conventional top-1-centric updates often reinforce errors by corrupting the local semantic geometry among related classes, while iterative adaptation exacerbates progressive bias accumulation, ultimately driving the model toward mode collapse. To overcome these coupled vulnerabilities, we propose Local Margin Restoration (LMR), a lightweight, one-step TTA framework. At the sample level, our Protected Margin Restoration (PMR) objective recovers local semantic geometry by shielding plausible near-top candidates from external hard negatives. Concurrently, to combat stream-level degradation, we introduce a dual-stage stabilization mechanism, featuring an Adaptive Margin (AM) controller and Bias Correction (BC), to dynamically disrupt progressive bias accumulation and prevent mode collapse. Extensive experiments on CIFAR-C, ImageNet-C, and ImageNet variants demonstrate that LMR consistently outperforms state-of-the-art TTA baselines, proving exceptionally robust and efficient even in challenging low-batch test-time regimes. Our code is available at https://github.com/DennisHuangYan/LMR.
VC-Tooler: Learning Compositional and Adaptive Visual Tool Use
VC-Tooler：学习组合性和自适应的视觉工具使用
Wu, Yizheng, Hua, Jiashen, Deng, Bing, Ye, Jieping
Abstract
Agentic multimodal reasoning extends passive image understanding by allowing VLMs to actively acquire and refine visual evidence through visual tool interactions. Effective visual tool use requires three capabilities: grounding tool calls in visual context, composing tools across multiple steps, and adapting reasoning to tool-returned observations. However, existing approaches largely focus on grounding within fixed tool spaces and rigid invocation patterns, leaving composition and adaptation insufficiently addressed. We present VC-Tooler, which learns visual tool use as a compositional and adaptive capability. To this end, we first build a trajectory bank through a hierarchical synthesis pipeline covering three capability levels: single-tool grounding, multi-tool composition, and diverse tool contexts and interfaces. We then train the model in two stages: a supervised cold start that establishes these capabilities, followed by reinforcement learning that encourages accurate, efficient, and context-aware visual tool use. VC-Tooler achieves state-of-the-art performance among open-source models on both general-purpose and agentic benchmarks, including $95.8\%$ on V* and $35.3\%$ on VTC-Bench, and shows promising transfer under richer tool settings at inference time. Project page: https://w1zheng.github.io/VC-Tooler
GenPrior: Unleashing Text-to-Motion Generative Priors for Zero-Shot Skeleton-based Action Recognition
GenPrior：释放文本到运动生成先验以实现零样本骨架动作识别
Kuang, Jidong, Wang, Hongsong, Gui, Jie
Abstract
Zero-shot skeleton-based action recognition (ZSAR) aims to recognize unseen action categories by aligning skeleton features with textual semantics. However, existing methods rely on text-derived prototypes that inherently lack geometric structure and physical constraints, resulting in a pronounced \textit{semantic-kinematic gap}. To bridge this gap, we propose \textbf{GenPrior}, the first framework to exploit generative priors from pre-trained Text-to-Motion (T2M) models for ZSAR. Specifically, we introduce Dispersion-Gated Feature Fusion, which distills kinematic prototypes and intra-class dispersion from generative motion sequences and employs a learned gating network to adaptively inject reliable structural cues into textual embeddings while suppressing synthetic artifacts. Furthermore, we propose Generative Prototype Refinement, which leverages these generation-enhanced prototypes as anchors to mine high-confidence unseen samples, calibrating class prototypes toward the true distribution and thereby unleashing strong performance gains. Extensive experiments on NTU-60, NTU-120, and PKU-MMD demonstrate that GenPrior achieves state-of-the-art performance under both zero-shot and generalized zero-shot settings. Code is available at https://github.com/jidongkuang/GenPrior.
Recent vision-language models for chest X-ray understanding are largely built on image-report alignment and therefore rely heavily on MIMIC-CXR as the dominant pretraining source. While effective at scale, this paradigm underexplores an important alternative source of supervision: a range of existing multi-label classification datasets, which provide cleaner and more explicit disease signals than free-text reports, and can offer broader pathology coverage when combined across sources. However, learning from such heterogeneous datasets is nontrivial, as differences in label ontologies, annotation protocols, acquisition pipelines, and report styles can cause models to entangle clinical semantics with dataset identity, leading to poor transfer despite increased scale. In this work, we revisit radiology VLM construction from the perspective of harmonized multi-source learning. We propose HarMoE, a dataset-aware mixture-of-experts framework that learns shared cross-dataset medical semantics while confining source-specific variation to lightweight residual experts in deeper decoder layers. To further exploit clean supervision from labeled datasets, we train in a unified disease vocabulary with masked multi-dataset supervision, enabling the model to leverage complementary annotations without introducing false negatives. Experiments on large-scale chest X-ray benchmarks show that HarMoE consistently improves zero-shot classification, out-of-distribution transfer, and grounding over strong baselines. Our results suggest that building robust radiology VLMs requires moving beyond single-source image-report alignment toward structured knowledge construction from heterogeneous datasets with cleaner supervision and broader coverage. Code and the 873k harmonized dataset will be released at https://github.com/Roypic/harmoe.
Rapidly evolving Generative AI enables sophisticated visual text manipulations that increasingly evade current forensic detectors. Existing discriminative models often overfit specific forgery patterns, limiting their generalization to unseen, open-set attacks. To address this challenge, we propose a generative detector that localizes tampering by estimating the local restoration cost required to align a query image with authentic visual-text statistics, rather than by learning forgery-specific decision boundaries. Specifically, we introduce Sparse-Constraint Rectified Flow (SC-RF), a detector-oriented adaptation of Flow Matching for spatially sparse anomaly localization. We further mitigate data scarcity via self-supervised Artifact Injection and preserve high-frequency forensic traces using a pixel-space Forensic-DiT. Extensive experiments on three benchmarks show that our method achieves state-of-the-art performance, surpassing the runner-up by 3.2 and 4.8 percentage points in F1 and IoU, respectively. In particular, the proposed detector demonstrates strong zero-shot performance on challenging unseen text editing patterns. We further provide an auxiliary stress-test analysis showing that local harmonization produced by our model can weaken the statistical cues relied upon by existing detectors, offering a complementary vulnerability-analysis perspective.
Open-vocabulary segmentation identifies and segments objects from arbitrary textual descriptions. SAM 3 supports noun-phrase-guided segmentation and achieves competitive open-vocabulary performance through exhaustive vocabulary traversal, yet suffers from prohibitive computational overhead as target categories scale. In this paper, we propose an Efficient Open-Vocabulary segmentation framework with SAM 3 (EOVSAM), which adapts SAM 3 for single-pass prediction. EOVSAM removes prompt conditioning to turn SAM 3 into an efficient mask generator and introduces a new Attentional Aggregation strategy to optimize open-vocabulary classification end-to-end. This formulation avoids the multi-stage pipelines and post-processing heuristics commonly used by existing methods, while mitigating the closed-set collapse that can arise when classification is optimized directly. EOVSAM consistently improves segmentation accuracy over vanilla SAM 3 on all evaluated datasets and accelerates inference by up to 338$\times$. Furthermore, EOVSAM maintains high accuracy at lower resolutions while achieving even more remarkable inference speeds. Experiments on standard semantic and panoptic segmentation benchmarks show that EOVSAM combines competitive or state-of-the-art accuracy with a substantial speed advantage over existing open-vocabulary segmentation models. Code and models are available at https://github.com/hustvl/EOVSAM.
We propose Sen-Cap, a Sensor-Flexible and Noise-Resilient 3D human motion Capture framework that integrates multi-modal data from LiDAR and camera. While multi-modal sensors provide richer information than single-modal sensors, existing approaches still suffer from two core challenges. First, multi-modal alignment/matching across arbitrarily deployed sensors is typically handled by explicit calibration, which propagates errors under changing viewpoints and in turn constrains deployment to fixed, highly overlapped layouts. Second, prior methods degrade under severe noise or partial sensor failures, which are common in real-world environments. To address these challenges, Sen-Cap introduces a Unified Across-Sensor Motion Estimator that reconstructs local pose and shape in a human-centric space without calibrations between sensors, supporting a flexible number of sensors, as well as a Noise-Resistant Trajectory Tracker that maintains robustness under severe point cloud noise through iterative refinement. These sensor-flexible and noise-resilient features make Sen-Cap more practical in real-world deployment. Notably, operating in real time, Sen-Cap achieves state-of-the-art performance on major metrics on Human-M3 and FreeMotion, as well as strong cross-domain performance on LiDARHuman26M and RELI11D. This combination of flexibility and robustness opens new opportunities for motion capture in real-world scenarios, e.g. sports analytics, field robotics, and large-scale immersive environments.
Extended Field of View Analysis for VideoGAN-based Trajectory Generation
基于VideoGAN的轨迹生成的扩展视野分析
Mariani, Annajoyce, Maag, Kira, Gottschalk, Hanno
Abstract
Realistic and diverse trajectory generation is central to enabling higher levels of vehicle automation. While rule-based and classical learning-based methods may struggle to capture the complexity of traffic behavior, generative models have already demonstrated in other fields that they can handle a comparable level of complexity. In this paper, we build upon previous work on generative adversarial network (GAN)-based semantic bird's-eye-view traffic generation and extend the proposed framework in several key aspects. We improve the semantic representation, replace the trajectory extraction procedure with a graph-based association method, and systematically investigate increasingly larger fields of view. In addition, we introduce a quantitative evaluation framework to assess hallucinations and object permanence in generated videos. Our experiments demonstrate that the framework generalizes to larger and more complex traffic scenes while maintaining statistically realistic trajectories and coherent spatial relationships between traffic participants. Within 150GPU hours of training and with inference times below 20ms for scenes of up to 20s, our results demonstrate that video-based GANs remain an efficient and scalable approach for realistic trajectory generation, even in substantially larger traffic scenes, making them well suited for downstream tasks such as prediction, planning, and simulation in automated driving.
ANN-based All-in-One image restoration (AiOIR) unifies diverse degradation handling but incurs high computational costs, limiting its real-time deployment. While Spiking Neural Networks (SNNs) offer a low-power alternative, applying them to static images remains challenging. This difficulty arises because explicit event signals are absent, and degradation cues are heavily entangled with scene structures, hindering the learning of reliable restoration-oriented spike events. To address these issues, we propose SpikeRestormer, an energy-efficient SNN for AiOIR that performs event reasoning over internally generated spike cues. Specifically, we propose a degradation-event perception process to extract spike-based degradation events through Subtractive Degradation Event Attention (SDEA). Moreover, we introduce Hierarchical Bayesian Skip Masking (HBSM) and Additive Restoration Event Attention (AREA) processes for event-reliability inference and restoration-event construction, respectively. By integrating these complementary processes, SpikeRestormer formulates restoration as a unified process of degradation-event perception, degradation-event reliability inference, and restoration-event construction, liberating the potential of SNNs for energy-efficient AiOIR. Extensive experiments show that SpikeRestormer delivers competitive performance against ANN-based methods and establishes new state-of-the-art results among SNN-based methods with significantly lower energy consumption.
Existing astronomy foundation models provide strong galaxy representations, but adapting them to new survey conditions and survey-specific morphology recognition tasks still requires substantial human supervision. We show that VLM-based VQA systems contain meaningful visual-semantic priors that can serve as weak supervision for downstream morphology classifiers and improve morphology classification under limited human-label budgets. We first introduce a survey-oriented VQA benchmark spanning two representative imaging regimes and evaluate state-of-the-art VLMs on galaxy morphology questions. The results show that these models capture useful morphology signals and informative uncertainty, but are not sufficiently reliable to replace human annotators. Motivated by this finding, we use a general-purpose VLM as a morphology teacher for Zoobot, an astronomy foundation model pretrained on large-scale Galaxy Zoo annotations. Across two survey domains and multiple annotation budgets, the VLM teacher consistently improves Zoobot's downstream morphology classification. These results demonstrate that a general-purpose VLM provides knowledge complementary to an astronomy foundation model and can teach it to better recognize galaxy morphology under limited human supervision. The resulting pipeline is designed for label-efficient adaptation to forthcoming large-scale surveys, including the Vera C. Rubin Observatory's Legacy Survey of Space and Time (LSST) and the Nancy Grace Roman Space Telescope. The benchmark and code are publicly available at https://github.com/fw-ic/VLM-morphology-teacher.
The Push-Forward Transform for Continuous and Robust Comparison of Dynamic Shapes
用于动态形状连续和稳健比较的推前变换
Rouatbi, Roua, Cardona, Juan-Esteban Suarez, Sbalzarini, Ivo F.
Abstract
We introduce a mathematical framework for shape comparison based on mapping functions from the shape domain to a common reference domain. This Push-Forward Transform enables invariant and robust comparison of shapes, preserving intrinsic geometric information. Quantitatively comparing shapes and their temporal evolution is a fundamental challenge in image analysis. Meaningful shape comparison requires representations that are invariant to transformations that do not alter shape itself, such as translation, rotation, reflection, re-parametrization, and uniform scaling, while remaining sensitive to intrinsic geometric variation. Existing approaches often rely on sensitive parameterizations, landmark correspondence, or learned representations that are difficult to interpret and reproduce. We show that the Push-Forward Transform (PF-T) applied to Signed Distance Functions (SDFs) yields a continuous representation that captures both boundary and interior geometry. We derive an interpretable morphometric that quantifies shape similarity and reveals features such as skeletal topology and rotational symmetries. The push-forward transform applies consistently to two- and three-dimensional shapes, extends to time-evolving geometries, and supports the joint analysis of shape and additional scalar fields defined over shapes, such as intensity or molecular signals. We present the mathematical formulation, describe an efficient algorithm, and benchmark the approach on 2D, 3D, and temporal data sets.
We present CalibBEV, a novel Bird's Eye View (BEV) alignment approach for LiDAR-camera calibration. Our method unifies LiDAR and camera data into a shared 3D spatial representation, enabling accurate and robust cross-modal calibration. CalibBEV extracts sensor-wise BEV features from each modality using domain-specific architectures and estimates the calibration matrix through a two-step alignment process. First, we perform an implicit alignment by regressing a coarse calibration matrix directly from the BEV features. To ease this alignment, we enforce semantic consistency between BEV representations across modalities using a contrastive loss inspired by CLIP, guiding both networks toward a unified feature space. In the second step, we leverage our BEV formulation to explicitly align the features of one modality with the other, refining the initial coarse estimate into a final, more accurate calibration matrix. CalibBEV significantly outperforms prior point-to-pixel matching methods, achieving state-of-the-art calibration accuracy. On the KITTI and nuScenes benchmarks, our method reduces the Relative Rotation Error (RRE) by 51% and 68%, and the Relative Translation Error (RTE) by 80% and 91%, respectively, compared to previous methods.
Geospatial foundation models aim to learn representations that transfer across regions and sensors, yet evaluating them on specific tasks requires large, high-quality, multi-modal benchmarks that measure how well such models extract value from data. Concerning flood mapping, existing datasets rarely combine bi-temporal SAR and co-registered optical imagery at scale, leaving the value of foundation models for this downstream task largely untested. We introduce GEOID-Flood, a large-scale multi-modal flood segmentation benchmark, derived from Copernicus Emergency Management Service activations, spanning 219 events across 65 countries over ten years. The dataset provides more than 14,000 tiles with co-registered pre- and post-event Sentinel-1, in GRD and RTC format, pre-event Sentinel-2 composite, and DEM, including manually validated labels that separate background from permanent water and flooded water. Using this benchmark, we evaluate foundation models against conventional encoders across single-image, multi-temporal, and multi-modal protocols. We report three main findings: foundation models offer a consistent but modest advantage; optical-SAR fusion with finetuning best resolves transient flooding; and models trained on GEOID-Flood transfer to unseen events better than those trained on existing datasets. Dataset and code available at https://github.com/links-ads/geoid-flood.
Accurate and efficient reconstruction of cloud-contaminated and noise-corrupted NDVI time series remains a challenge in remote sensing. Deep learning provides a promising solution for modeling complex spatiotemporal dependencies; however, its application is often limited by the difficulty of obtaining paired clear-sky and degraded NDVI data for identical spatiotemporal locations. To address this issue, we propose GloSSR, a Global-scale Self-supervised Spatiotemporal framework for NDVI Reconstruction. The framework constructs supervisory signals by artificially degrading relatively clean NDVI observations with realistic cloud contamination patterns, producing self-supervised training pairs that closely mimic real-world degradation. It further introduces an end-to-end spatiotemporal learning network that jointly captures long-range temporal dependencies and short-term spatiotemporal correlation through a bidirectional Transformer with a ConvLSTM architecture. A temporal-channel attention-based reconstruction module is incorporated to enhance informative features, while a spatiotemporal prior constraint is designed to preserve both fine-scale structures and long-term phenological trends during optimization. Extensive evaluations on MODIS NDVI data demonstrate the effectiveness of the proposed framework across both artificial and real-world scenarios. In artificial degraded-pixel reconstruction experiments, GloSSR consistently outperforms the comparison methods. Time-series analyses based on real observations further demonstrate that the proposed framework can accurately characterize vegetation dynamics and capture the key phenological states. Long-term vegetation trend analysis and the transferability analysis to AVHRR data validate the scalability of the framework and illustrate its broad applicability for large-scale environmental monitoring.
Implicit Neural Representations for Multimodal Longitudinal Image Imputation and Interpolation
用于多模态纵向图像插补和插值的隐式神经表示
Wendrich, Sina, Förner, Lukas, Reinke, Zoe, Tehlan, Kartikay, Berlis, Ansgar, Frühwald, Michael, Wagner, Matthias, Wendler, Thomas
Abstract
Longitudinal multiparametric MRI is central to follow-up imaging in oncology, yet real-world clinical data are characterised by missing sequences, heterogeneous acquisition protocols, and varying spatial resolutions across time points. We propose a patient-specific conditional implicit neural representation (INR) that models multimodal longitudinal MRI as a continuous function of world coordinates, time, and modality conditioning. The model is trained with stochastic modality dropout to handle incomplete data, and its continuous coordinate-space formulation enables both spatial and temporal interpolation without resampling to a fixed voxel grid. A self-consistency-based confidence estimator is derived from cross-modal reconstruction performance at inference time. We evaluate the framework on longitudinal MRI from paediatric brain tumour patients, demonstrating statistically significant improvements over linear interpolation for T1CE and FLAIR (p < 0.05), with mean MS-SSIM of 0.95 $\pm$ 0.02 for T1CE. Predicted confidence correlates strongly with true reconstruction quality (Pearson r up to 0.996), suggesting reliable deployment potential in heterogeneous clinical settings.
Context-Aware Mixture of Domain Experts for Bodily Expression of Emotion in the Wild
面向上下文的领域专家混合模型用于野外情感的身体表达
Dehshibi, Mohammad Mahdi, Masip, David
Abstract
The same body posture can convey entirely different emotions depending on its surrounding context, yet most methods for recognising bodily emotions treat scene and object cues as auxiliary feature augmentations rather than as structured priors over the plausibility of emotions. We introduce the Context-Aware Mixture of Domain Experts (CA-MoDE) for bodily emotion recognition. CA-MoDE incorporates dedicated scene and object experts to generate soft distributions over emotion categories conditioned on their respective domains. These domain-conditioned soft predictions serve as structured contextual priors that modulate the body expert's predictions at the distributional level rather than at the feature level. To fuse these multi-domain signals, we propose a task-tailored max-endorsement gating strategy that selects the strongest contextual signal across experts for each emotion dimension. Our gating strategy mitigates the signal dilution that typically occurs when conflicting or uninformative context distributions are averaged. CA-MoDE achieves an Emotion Recognition Score of 0.3269 on the Body Language Database. By outperforming existing temporal models using only single still images, our framework demonstrates that explicitly modelling structured spatial context can serve as a complementary discriminative proxy for the behavioural dynamics typically captured by video.
Chinese Translation
相同的身体姿势可以根据其周围上下文传达完全不同的情感，然而大多数识别身体情感的方法将场景和物体线索视为辅助特征增强，而不是作为情感可能性的结构性先验。我们提出了面向上下文的领域专家混合模型（Context-Aware Mixture of Domain Experts, CA-MoDE）用于身体情感识别。CA-MoDE结合了专门的场景和物体专家，以生成基于各自领域的情感类别的软分布。这些领域条件下的软预测作为结构性上下文先验，调节身体专家在分布层面而非特征层面的预测。为了融合这些多领域信号，我们提出了一种针对任务的最大认可门控策略，该策略为每个情感维度选择最强的上下文信号。我们的门控策略减轻了在冲突或无信息的上下文分布被平均时通常发生的信号稀释现象。CA-MoDE在身体语言数据库上实现了0.3269的情感识别评分。通过仅使用单张静态图像超越现有的时间模型，我们的框架表明，明确建模结构化空间上下文可以作为通常由视频捕捉的行为动态的补充判别代理。
Old photographs often suffer from multiple coupled degradations, including scratches, cracks, fading, blur, noise, and missing regions, severely degrading both visual quality and semantic content. We propose Loop-Mamba, a lightweight loop-based state-space framework that formulates old photo restoration as progressive state evolution, where a persis- tent restoration state is continuously propagated and refined through iterative computation. Specifically, we introduce a Semantic-Guided Degradation Estimator (SGDE) to explicitly model heterogeneous degradations by jointly predicting local degradation maps and global degradation scores, providing degradation-aware guidance for state evolution. We further develop a Shared Structural Memory Mamba (S$^2$M- Mamba), which maintains a persistent restoration state across iterations, enabling persistent state evolution through shared structural memory for robust long-range structural reconstruction. Benefiting from first-order state recursion, Loop-Mamba propagates latent restoration states through recurrent tran- sitions instead of repeatedly stacking deep feature transformations, thereby alleviating gradient dilution while avoiding the computational overhead inherent in iterative CNN- and Transformer-based restoration frameworks. A lightweight multi-directional scanning strategy further enhances direc- tional information aggregation and preserves structural continuity. To better evaluate restoration quality, we introduce the task-oriented Old Photo Damage Recovery Score (ODRS), which jointly measures degradation recovery and structural reconstruction fidelity. Experimental results on the public SynOld benchmark demonstrate that Loop-Mamba consistently outperforms previous state-of-the-art methods across both conventional restoration metrics and the proposed ODRS.
A wearable assistant should both answer questions about its visual history and recognize when that history is useful to the present situation. Existing video-memory systems primarily support question-conditioned recall, whereas proactive assistants typically use separate memory and control mechanisms. We introduce GROVE, a training-free framework that supports both behaviors with one memory grown causally from a continuous video stream. GROVE retains fine-grained perceptual evidence and incrementally consolidates it into time-stamped moments, coherent episodes, and recurring cross-day patterns. Each stratum is paired with a scale-native retrieval skill for locating an observation, replaying an activity, or traversing long-range regularities. Reactive QA and proactive assistance share this memory and access interface, differing in whether retrieval is initiated by a user query or the current situation. Across multiple benchmarks including the challenging MM-lifelong and EgoServe, GROVE achieves the best results among the compared methods. Controlled ablations show that the temporal strata and their access skills are complementary, with patterns providing the largest benefit when evidence spans multiple days. Code will be available at https://github.com/SitongGong/GROVE.
Most evidence on the effectiveness of explainable artificial intelligence (XAI) attribution methods has been established on convolutional neural networks (CNNs), with limited investigation into whether these conclusions generalize to the diverse Vision Transformer (ViT) architectures that now dominate computer vision. This paper presents a controlled benchmark that evaluates attribution quality across five dimensions: faithfulness, localization, robustness, complexity, and computational cost. A standardized framework assesses 13 attribution methods from four algorithmic families on eight representative backbones spanning CNNs, isotropic ViTs, hierarchical transformers, hybrid architectures, and linear-attention transformers. The results show that attribution performance is strongly architecture-dependent and that rankings established on CNNs do not reliably transfer to transformer-based models. CAM-based methods achieve the highest scores under the conventional bounding-box localization metric on CNNs and most ViTs but perform poorly on linear-attention architectures. Pixel-level dense-mask evaluation further reveals that these gains largely reflect metric saturation rather than accurate localization. CAM-based methods also exhibit limited robustness on global-attention transformers, whereas attention rollout provides consistently stable explanations with poor localization. Furthermore, faithfulness correlation offers limited discrimination between attribution methods, highlighting the limitations of single-metric evaluation. These findings challenge prevailing conclusions on attribution performance and demonstrate the need for architecture-aware, multi-dimensional evaluation. The open-source code for the evaluation framework and benchmark results is available at https://github.com/Nishan-Charlie/VIT_XAI_Bench.
USP-Mamba: Unmixing-Derived Spectral and Structural Prompting for Hyperspectral Image Super-Resolution
USP-Mamba：基于解混合的光谱与结构提示用于高光谱图像超分辨率
Chen, Shi, Zhang, Jie, Zhou, Yicong
Abstract
Hyperspectral image super-resolution aims to reconstruct high-resolution imagery while preserving dense spectral information. Recently, Mamba-based models have shown promising potential for this task by capturing long-range dependencies with linear computational complexity. Nevertheless, their causal sequence modeling requires two-dimensional hyperspectral features to be unfolded along predefined scanning orders, which disrupts spatial adjacency and restricts the effective propagation of contextual information. Moreover, state-space parameterization of existing models is predominantly derived from generic learned representations, without explicit alignment with the intrinsic characteristics of the hyperspectral image. To address this issue, we propose an Unmixing-derived Spectral and Structural Prompting Mamba framework, termed USP-Mamba, which adapts Mamba state evolution through composition-aware spectral priors and image-dependent structural prompts. Specifically, an unmixing-informed spectral prompt captures the global material composition of the input image and provides persistent conditioning throughout reconstruction. Injected into the Mamba sequence and progressively adapted across layers, it steers state evolution toward composition-consistent reconstruction. We introduce feature-level structural prompts comprising spatial and frequency components to provide image-dependent local guidance. The spatial prompt promotes structure-sensitive state encoding for local detail preservation, while the frequency prompt enables region-adaptive transitions between homogeneous regions and high-frequency details. Finally, complementary Hilbert and Semantic-Guided Neighboring scans preserve spatial continuity and strengthen non-local semantic dependency modeling. Extensive experiments on different datasets demonstrate that the proposed method consistently outperforms representative approaches.
The Loggia dei Lanzi in the Piazza della Signoria is one of Florence's most prominent structures visited by millions every year. Its construction history spans multiple centuries of modification. This paper presents the results of a thermal imaging campaign conducted in December 2025, using a FLIR T1020 HD camera, revealing hidden architectural features including walled-up openings and material transitions beneath the plaster surface. The favorable winter ambient conditions provided a feature-rich benchmark upon which to compare the results of enhancement algorithms and artificial intelligence models. We evaluate the application of AI-based image enhancement to thermal heritage documentation through a comparison of three tiers of image resolution in a photogrammetric Structure-from-Motion (SfM) pipeline: native resolution, FLIR's hardware-based pixel-shifted super-resolution (UltraMax), and state of the art AI-upscaled imagery models. We quantify the effect of each resolution tier on feature detection and tie-point generation, assessing whether the additional detail produced by super-resolution, whether hardware or AI-derived, translates into meaningfully denser and more accurate 3D thermal models. Our results contribute to the emerging intersection of artificial intelligence and heritage thermography by providing a direct comparison of hardware microscanning and AI super-resolution within a thermal photogrammetric workflow for cultural heritage. All datasets are made publicly available and accessible within an interactive 3D archival framework, and integrated into a custom citywide extended reality overlay application.
Forecasting future states from video sequences is a critical challenge for autonomous robotic systems and a fundamental objective of world modeling. Prior generative methods operating at the pixel level inevitably overemphasize task-irrelevant details, leading to prohibitive computational overhead. While latent-based approaches attempt to mitigate this by predicting features directly, the persistent reliance on heavy decoders for state-to-task mapping remains a computational bottleneck. In this work, we propose Decoder-Free Feature Forecasting (DF$^3$), a novel framework that models world evolution entirely within the latent space and directly derives task outputs, completely eliminating the need for a decoder. Specifically, DF$^3$ injects learnable spatial queries into the terminal blocks of a frozen vision foundation model to extract future state representations directly. By employing a lightweight, unified Motion-Aware Context Fusion (MACF) mechanism that seamlessly integrates coarse flow warping with fine-grained latent cross-correlation, these queries interact with historical token representations to explicitly align and forecast the feature of the next frame. Subsequently, a specialized set of task queries probes these forecasted features for the downstream task. Extensive experiments on public benchmarks and zero-shot deployment in a robotic simulator demonstrate that DF$^3$ achieves performance comparable to state-of-the-art methods while offering superior efficiency and flexibility for integrated perception and control.
Autoregressive models have emerged as an effective paradigm for point cloud generation. However, most existing approaches rely on heuristic tokenization strategies, such as spatial sorting or stochastic downsampling, which often disrupt intrinsic point cloud topology and weaken the structural coherence of the generated shapes. In this paper, we present PointRSP, an autoregressive framework that reformulates point cloud generation as a topology-preserving tessellation process via recursive spectral partitioning. Instead of constructing token sequences heuristically, we introduce a topology-aware partitioning autoencoder that decomposes an unstructured point cloud into a non-balanced binary tree through a hybrid recursive spectral partitioning strategy. This hierarchical representation provides a deterministic geometric blueprint that preserves topological relationships while capturing multiscale structural dependencies within a quantized latent space. To synthesize shapes in this space, we propose a dual-stream cascaded generator that jointly models structural evolution and feature synthesis. In addition, we design a geometry-calibrated positional encoding mechanism that anchors latent embeddings using multi-scale structural centers, which stabilizes cascaded generation during the early stages of structural formation. Extensive experiments show that PointRSP achieves state-of-the-art performance in generation quality and diversity, demonstrating strong generalization across complex 3D topologies.
Single-image feed-forward 3D Gaussian Splatting (3DGS) aims to directly generate a renderable 3D scene representation from one input image, avoiding the cost of multi-view capture and per-scene optimization. However, existing methods are often constrained by a pixel-aligned representation, where Gaussians are predicted from fixed image-grid locations. Such pixel-aligned primitives can produce promising nearby-view renderings, but they remain weakly coupled to underlying scene surfaces and struggle to preserve coherent structures under large viewpoint shifts. We present InfiniSplat, a feed-forward single-image 3DGS framework that moves from a pixel-aligned representation toward a surface-aligned representation. InfiniSplat constructs this representation by first using geometry-guided sampling to place 2D supports according to depth-induced local surface structure, and then applying a query-conditioned implicit decoder to predict Gaussian attributes from the image features queried at these supports.By grounding support locations in geometry while decoupling Gaussian prediction from fixed pixel centers, InfiniSplat produces Gaussian layouts that better follow scene surfaces and reduce scattered primitives caused by grid discretization.Across multiple cross-dataset NVS evaluations, InfiniSplat achieves state-of-the-art performance compared with single-image feed-forward baselines, and demonstrates zero-shot generalization from Hypersim indoor synthetic training to complex open-world scenes.Project page: https://zju3dv.github.io/InfiniSplat.
Rip currents are recurrent coastal natural hazards that threaten beachgoers and create operational challenges for lifeguards and coastal managers. Reliable monitoring from standard RGB (red-green-blue) imagery acquired by unmanned aerial vehicles (UAVs) remains difficult because hazardous channels often appear as subtle gaps in breaking waves, foam texture, or sediment patterns, and these signatures are affected by illumination, sea state, and environmental noise. This study presents a physically informed coastal environmental monitoring workflow for detecting visually expressed rip-current indicators that integrates wavelet-derived spatial-frequency texture features with deep learning. We evaluate multiple strategies for incorporating Discrete Wavelet Transform features into convolutional architectures, from computationally efficient channel replacement to dual-stream fusion with attention mechanisms. Performance is assessed against a standard RGB baseline using a task specific convolutional neural network for image-level presence classification and a YOLOv8 model for object-level localization. Under the evaluated dataset conditions, integrating wavelet derived texture features improves performance over RGB-only models. The dual-stream architecture achieves the strongest classification performance, exceeding 95% accuracy with high recall, while channel replacement is most effective for YOLOv8 object detection, reaching 94% mAP@50 for localization. Explainable artificial intelligence analyses provide qualitative evidence that the models attend to visually plausible wave-gap regions associated with rip currents. These results suggest that under the conditions of the evaluated dataset, physically informed wavelet integration may support UAV-based decision-support tools for interpretable beach-safety risk mitigation.
Deploying vision-language models (VLMs) for safety-critical spatial reasoning on resource-constrained autonomous driving platforms requires both compact model size and reliable metric grounding. We present MoRAL (Multimodal Reasoning for Autonomous Language Models), a two-stage fine-tuning pipeline that teaches Cosmos-Reason2-2B to first read a physics-encoded Bird's Eye View (BEV) representation and then reason over it for driving decisions. The BEV image encodes LiDAR metric distance as color bands, object class as cluster morphology, and radar Doppler velocity as directional wedge overlays, externalizing spatial perception into the input image so that no learned 3D backbone is required at inference. Stage 1 fine-tunes the vision encoder on 60,000 grounding records; zero-shot baselines produce no parseable BEV outputs, confirming the vocabulary requires explicit training. Stage 2 fine-tunes the full model (52M parameters, 2.4% of total) on 57,696 chain-of-thought records generated by Cosmos-Reason2-8B as teacher, spanning eight driving question types. On 2,304 held-out nuScenes frames evaluated by Gemma 4 (31B) calibrated against human review, MoRAL wins seven of eight question types over a zero-shot 8B baseline despite using four times fewer parameters, with the largest margins on question types requiring structured multi-step physics reasoning. Emergency braking recall improves from 10.8% to 47.8%, output degeneration falls from 94.1% to 20.8%, and the full pipeline fits a consumer 8 GB GPU at 42 tok/s without quantization. These results establish a reproducible foundation for compact, physics-grounded VLM reasoning on mobile edge platforms.
Interactive segmentation reduces the prohibitive cost of pixel-level annotation by allowing users to delineate objects with a few clicks. However, applying this paradigm directly to remote sensing imagery is non-trivial: ultra-high resolutions, small object sizes, and sparse spatial distributions all degrade segmentation quality. Recent work has addressed the resolution barrier and achieved competitive results in interactive segmentation for remote sensing (ISRS). However, they treat all instances of a class within an image as a single objective target. Consequently, interactions spent on one object contribute nothing to its same-class neighbours, and satisfactory masks may demand up to 40 clicks per image, hindering the practicality of these frameworks. We observe that remote sensing scenes exhibit markedly strong inter-object correlation, meaning a single clicked object is highly informative about the rest of its category. Building on this, we propose ISRS-DETR, a detection-guided interactive segmentation framework that injects object-level evidence into both training and inference. Our ISRS-DETR employs an RF-DETR decoder with the interactive segmentation backbone to localise co-occurring same-class objects, and introduces a Dynamic Top-K Click Selection strategy that retains only reliable proposals and converts each into a simulated click, so one user interaction propagates across an entire class. Experiments on three standard remote sensing benchmarks show that ISRS-DETR achieves state-of-the-art accuracy while substantially reducing Number of Clicks per Image (NoC-I). All codes and data splits will be released for reproducibility upon acceptance.
AnimalCLEF26 addresses discovery-oriented animal re-identification, where systems must both attach query images to known individuals and discover unseen individuals by clustering them correctly. We present a similarity-to-clustering pipeline for this setting across Eurasian lynx, fire salamander, loggerhead sea turtle, and Texas horned lizard images. The method first isolates the target specimen using segmentation and then applies lightweight species-specific preprocessing for lynx, sea turtle, and salamander images to enhance identity-relevant visual cues, while Texas horned lizard images are used after segmentation only. Pairwise similarities are then estimated with WildFusion by calibrating and combining a MiewID global descriptor with two local matching branches, ALIKED + LightGlue and DISK + LightGlue. The resulting query-query similarities are refined and converted into identity clusters using graph-based clustering, while query-database similarities are used to attach confident samples to known identities. We evaluate training-free and fine-tuned MiewID variants, including Dynamic ArcFace and SphereFace2-Focal adaptations, and combine them in the final ensemble. Our selected ensemble substantially improves on the WildFusion baseline, achieving the best public ARI of 0.72124 and a private ARI of 0.70393, while a simpler preprocessing-before-calibration variant achieves the best private ARI of 0.71087. These results indicate that calibrated global-local fusion with species-aware preprocessing choices is effective for open-set wildlife re-identification under challenging field conditions and visual variation. The implementation code is available on GitHub.
Chinese Translation
AnimalCLEF26 关注以发现为导向的动物重识别任务，在该任务中，系统不仅需要将查询图像与已知个体关联，还需通过正确的聚类发现未见个体。我们提出了一种适用于这一场景的相似性到聚类的流程，涉及欧亚猞猁、火蝾螈、红海龟和德克萨斯角蜥的图像。该方法首先通过分割技术孤立目标标本，然后对猞猁、海龟和蝾螈图像应用轻量级物种特定的预处理，以增强与身份相关的视觉线索，而德克萨斯角蜥图像仅在分割后使用。接着，通过校准和结合 MiewID 全局描述符与两个局部匹配分支（ALIKED + LightGlue 和 DISK + LightGlue），使用 WildFusion 估计成对相似性。生成的查询-查询相似性经过精细化处理并转化为身份聚类，而查询-数据库相似性则用于将可信样本附加到已知身份上。我们评估了无训练和微调的 MiewID 变体，包括动态 ArcFace 和 SphereFace2-Focal 适配，并将它们结合在最终的集成模型中。我们选择的集成模型在 WildFusion 基线之上显著提升，达到了最佳公共调整兰德指数（ARI）0.72124 和私有 ARI 0.70393，而一个更简单的预处理-后校准变体则达到了最佳私有 ARI 0.71087。这些结果表明，结合物种感知的预处理选择的校准全局-局部融合在具有挑战性的野外条件和视觉变化下，对于开放集野生动物重识别是有效的。实现代码可在 GitHub 上获取。
Vision-language models (VLMs) are increasingly deployed as reasoning agents in real-world visual assessment pipelines, yet their spatial grounding remains unreliable for fine-grained, visually ambiguous targets. We study this gap in the context of automated vehicle damage assessment, where fine-grained defects such as scratches and hairline cracks occupy few pixels, produce weak gradient signal, and are easily confused with reflections and surface texture. We show that a state-of-the-art VLM (Qwen-VL) achieves strong semantic classification accuracy (87.3%) on this task but is systematically ungrounded at the spatial level: it hallucinates damage in reflective regions, misses elongated scratches entirely, and produces spatially inconsistent outputs when prompted for localization. We propose TinyDamage, a hybrid architecture that delegates spatial grounding to a dedicated multi-task segmentation model while reserving the VLM for semantic reasoning and report generation. On the segmentation side, we find that the choice of loss function has an outsized and underexplored effect on tiny-object grounding: focal loss, widely used for class imbalance, collapses tiny-damage detection to zero, while a supervised contrastive objective measurably improves damage/background separability. We integrate the segmentation model into a 7-node LangGraph agent pipeline that grounds every VLM generation step in the segmentation output, and show that this grounding reduces the report hallucination rate from 92% (text-only) and 78% (image-only) to 31% in a controlled evaluation on 100 human-verified reports. We introduce DET_l, a permissive per-category detection metric for evaluating tiny-object grounding under class imbalance, and report latency and reliability characteristics of the deployed pipeline.
Computational attention models could help surgeons manage the visual demands of laparoscopy, but they require dense spatial labels that are difficult to obtain because surgical intent is highly specialized and tacit. Here, we introduce DiffeoAfford, an action-grounded tissue affordance framework that retrospectively derives visual attention supervision from completed surgical procedures. By combining diffeomorphism-constrained tissue tracking with instrument trajectory analysis, DiffeoAfford generates affordance hotspot labels without manual per-frame annotation. A real-time prediction model trained on these labels anticipates relevant surgical regions and enables AffordView, an assistive auto-framing system for laparoscopic visualization. The proposed framework aligns with expert annotations and intraoperative surgeon gaze, and reduces surgeon cognitive workload during real-world evaluations using subjective, physiological, and behavioral measures.
Audio-driven video generation (A2V) has achieved promising progress in synthesizing temporally coherent and audio-visually aligned videos, yet its inference remains expensive due to the iterative denoising process of diffusion models. Existing caching methods mainly exploit temporal redundancy in visual features while overlooking the cross-modal alignment of A2V, where audio drives visual generation with highly non-uniform temporal importance. In this paper, we identify two levels of misalignment in existing A2V caching methods: temporal-semantic and computation-storage misalignment. To address them, we propose EchoCache, an energy-guided cross-modal caching framework for efficient A2V generation. EchoCache leverages audio time-frequency energy as a saliency anchor to guide latent-level cache updates and further introduces a dynamic timestep-latent caching mechanism with quantized cache management for joint efficiency and memory optimization. Extensive experiments on mainstream A2V models show that EchoCache consistently improves the latency-quality trade-off while preserving generation quality and audio-visual consistency. In particular, on Wan2.2-S2V over the EMTD benchmark, EchoCache achieves a 2.46x speedup with the best overall performance. Code is available at https://github.com/IF-LAB-PKU/EchoCache.
Fermat Active Laplace Learning for Semi-Supervised Hyperspectral Image Classification
基于费马活跃拉普拉斯学习的半监督高光谱图像分类
Buranasiri, Vutichart, Murphy, James M.
Abstract
Two active learning algorithms for hyperspectral image (HSI) classification are proposed that combine density-aware Fermat distances with Poisson-reweighted harmonic label propagation. Our methods actively query points using an uncertainty-based acquisition function, extending Poisson ReWeighted Laplace Learning (PWLL). Our first algorithm, Fermat Active Laplace Learning (FALL), builds an affinity matrix using Fermat distances between all data points. Then, PWLL is run with a diagonal perturbation using the minimum-norm acquisition function. In contrast, Approximate FALL (A-FALL) computes Fermat distances between each data point and landmark pixels selected via farthest-point sampling and constructs the affinity matrix using landmark multidimensional scaling. After several query rounds, A-FALL selects the Fermat exponent $p$ using a leave-one-out cross-validation variant. FALL and A-FALL leverage Fermat distances and subsequent harmonic label propagation to provide a density-aware estimation of the data manifold, improving labeling accuracy. Experiments on Salinas A and Pavia show the effectiveness of FALL and the scalability of A-FALL to large HSI scenes.
DyFrDet: Towards Accurate Small Object Detection via Dynamic Frequency Suppression with Label Disambiguation
DyFrDet：通过动态频率抑制与标签消歧来实现准确的小物体检测
Yang, Zihan, Guo, Yang, Zhang, Hongxing, Lu, Dan, Yao, Siyuan
Abstract
Despite the remarkable progress over the past decades, accurately identifying small objects remains challenging because of their insufficient visual cues. Previous works typically attempt to construct discriminative representation of the small objects. However, the wide range frequency domain noises and label ambiguities have been greatly overlooked, which significantly hinders the accurate localization. To address these issues, we propose a novel small object detection (SOD) detector termed DyFrDet, which is able to precisely localize the small object by dynamically suppressing the background distractions in frequency domain. Specifically, we propose a Dynamic Frequency-aware Feature Pyramid Network (DyFrFPN) to adaptively suppress low-frequency redundancy and excessive high-frequency noises. The DyFrFPN transforms the hierarchical features into frequency domain representation, and introduces a Dynamic Band Predictor (DBP) to preserve the discriminative components for small object identification. Afterwards, we present a novel Label Disambiguation Module (LDM), which leverages probabilistic distributions to explicitly model and alleviate the inherent ambiguity of target labels, yielding efficient improvement in localization precision of the small objects with low-resolution. Extensive experiments demonstrate that DyFrDet achieves state-of-the-art performance across multiple benchmarks, indicating its effectiveness and robustness in various challenging scenarios. Our code is available at https://github.com/ManOfStory/DyFrDet.
Video Diffusion Transformers (VDiTs) enable high-fidelity generation but incur quadratic cost from dense 3D self-attention. Existing head- and block-level sparse methods share computation budgets across queries, overlooking token-specific attention demand. We observe that retained density varies across queries yet correlates log-linearly with attention entropy, while dominant interactions form query-centered neighborhoods with token-dependent radii. Based on these findings, we propose Token Radius Attention (TRA), a training-free framework that maps query entropy to an analytic token budget and converts it into a temporally decayed radius without explicit key ranking. Fused entropy extraction, warm-up reuse, and block-sparse mask construction further reduce overhead. Across seven Wan2.1, Wan2.2, and HunyuanVideo T2V/I2V configurations, TRA retains only 9-19% of attention interactions and achieves 1.56x-2.05x speedup with competitive generation quality. Code is available at https://github.com/IF-LAB-PKU/Token-Radius-Attention.
Automated pain assessment in real clinics is limited by scarce clinically grounded facial video data with weak labels (often sequence-level self-report) and by the fact that pain cues can be subtle or near-neutral in RGB, while thermal and depth signals are informative yet impractical to deploy routinely. To address these challenges, we propose ReMiX-MAE (Reconstructing Missing Channel Cross-Modal Masked Autoencoder), a self-supervised multimodal masked pretraining framework that learns transferable facial representations from synchronized RGB, thermal, and depth videos and explicitly trains robustness to missing modalities, enabling RGB-only deployment. To fill the gap of clinically grounded facial pain data with video-level self-report and longitudinal treatment trajectories, we collect the Sympathetic Mediated Pain (SMP) dataset with paired pre- and post-recordings across multiple visits. Under RGB-only deployment, we evaluate ReMiX-MAE using both direct feature extraction and pseudo-multimodal features decoded from RGB. ReMiX-MAE consistently outperforms an RGB-only masked autoencoder baseline on SMP, with pseudo-multimodal features providing additional gains in the challenging five-class setting. Across external datasets, ReMiX-MAE further shows more robust and label-efficient transfer than RGB-only baselines, highlighting its advantage in data-limited clinical settings.
UEmbed: Unified Sparse and Dense Multimodal Embeddings
UEmbed：统一稀疏与密集多模态嵌入
Song, Tingyu, Li, Mingxin, Zhang, Yanzhao, Long, Dingkun, Xie, Pengjun, Nie, Zhijie, Zhao, Yilun, Wu, Shu
Abstract
Sparse retrieval underpins modern search systems, from web search to retrieval-augmented generation. Existing work has introduced Learned Sparse Retrieval (LSR) to push beyond exact lexical matching toward richer semantics. Yet LSR has so far remained tied to encoder-style bidirectional architectures, and its extension to multimodal settings still relies heavily on auxiliary cross-modal modules. To address these limitations, we introduce UEmbed (Unified Embedding), a decoder-only multimodal embedding model that produces both sparse lexical and dense representations in one causal forward pass. UEmbed appends N learnable special tokens to the input and partitions the vocabulary into N disjoint subsets. Each token's causal hidden state predicts sparse weights over its assigned subset, and the N subsets are concatenated into the full sparse vector. Trained on public data, we release UEmbed at 2B, 4B, and 9B scales. UEmbed-9B reaches 71.8 (dense) and 71.0 (sparse) on MMEB-v2, outperforming multimodal embedding models trained on publicly available data (e.g., RzenEmbed). On BEIR, UEmbed also remains competitive with strong dense and sparse baselines. Furthermore, we demonstrate the practical utility of UEmbed across three dimensions: effectiveness, efficiency, and agentic applications. Overall, UEmbed offers a new paradigm: it unifies dense and sparse embeddings in one model, while further extending sparse retrieval to unify text and multimodal inputs.
CAPEval: A Decoupled Caption Evaluation across Understanding and Generation
CAPEval：一种解耦的图像理解与生成中的字幕评估
Liu, Zhipeng, Wang, Haochen, Zhang, Zhaoxiang
Abstract
Captions serve as a primary supervision signal for both multimodal understanding and text-to-image generation. However, previous evaluations treat the caption quality as a single scalar objective, which conflates two distinct properties: (1) how much visual information a caption covers and (2) how reliably the image supports its stated claims. To this end, we design a decoupled caption evaluation benchmark, CAPEval (Coverage And Precision Evaluation), with human-written ground-truth captions and human-verified atomic checklist items. Specifically, CAPEval decomposes caption quality into Coverage and Precision. The former quantifies how thoroughly a caption covers ground-truth factual content, while the latter reflects the factual correctness rate of all claims expressed in the caption. We select 10 captioners and further conduct controlled downstream end-to-end experiments with them from four model families, where the caption source is the only variable. Empirically, we find a consistent task-dependent dissociation: Coverage serves as the stronger correlate for understanding performance, whereas Precision acts as the dominant predictor for generation performance. This decoupled evaluation paradigm not only delivers a more fine-grained diagnosis of caption quality, but also offers actionable guidance for selecting and optimizing captioners tailored to different downstream tasks.
Chinese Translation
字幕作为多模态理解和文本到图像生成的主要监督信号。然而，以往的评估将字幕质量视为单一的标量目标，这混淆了两个不同的属性：（1）字幕覆盖了多少视觉信息，以及（2）图像在多大程度上支持其所陈述的主张。为此，我们设计了一个解耦的字幕评估基准，CAPEval（Coverage And Precision Evaluation），该基准包含人工撰写的真实字幕和经过人工验证的原子检查项。具体而言，CAPEval将字幕质量分解为覆盖度和精确度。前者量化字幕对真实事实内容的覆盖程度，而后者反映字幕中所有主张的事实正确率。我们选择了10个字幕生成器，并进一步对来自四个模型家族的控制下游端到端实验进行研究，其中字幕来源是唯一的变量。通过实证研究，我们发现了一种一致的任务依赖性解耦：覆盖度作为理解性能的更强相关因素，而精确度则是生成性能的主要预测因子。这种解耦的评估范式不仅提供了对字幕质量的更细致诊断，还为根据不同下游任务选择和优化字幕生成器提供了可操作的指导。
VR3D: View-Robust 3D Representation Learning for Aerial-Ground Person Re-Identification
VR3D：用于空中-地面行人重识别的视角鲁棒3D表示学习
Ji, Chao, Xuan, Shiyu, Li, Zechao
Abstract
Aerial-ground person re-identification is a challenging task due to cross-platform viewpoint variations, which cause severe occlusion and geometric deformation. Existing methods attempt to learn view-invariant representations exclusively within the 2D image space, where drastic viewpoint variations cause the learned features to remain coupled with viewpoint bias. To address this, we propose VR3D, a View-Robust 3D Representation Learning framework that maps images into a unified 3D coordinate space to achieve view-independent feature interaction. Specifically, we introduce View-Robust 3D Representation Interaction, which leverages 3D priors extracted from single 2D observations to lift 2D appearance features into a canonical 3D space. VR3I employs 3D Geometry-Semantic Attention to establish interactions between 2D patches and 3D voxels from corresponding body parts based on their 3D spatial locations, effectively grounding 2D semantics within a 3D framework. In addition, as the reliability of these representations varies across samples due to viewpoint changes and 3D reconstruction errors, we introduce Reliability-Aware Fusion, which estimates sample-specific reliability and adaptively aggregates the multi-source representations. Extensive experiments on three benchmark datasets (CARGO, AG-ReID.v1, and AG-ReID.v2) demonstrate that VR3D outperforms recent methods. For example, it achieves a 5.63% improvement in Rank-1 on CARGO. Our code will be released.
Chinese Translation
空中-地面行人重识别是一项具有挑战性的任务，主要由于跨平台视角变化导致严重的遮挡和几何变形。现有方法尝试在二维图像空间内学习视角不变的表示，但剧烈的视角变化使得学习到的特征与视角偏差保持耦合。为了解决这一问题，我们提出了VR3D，一个视角鲁棒的3D表示学习框架，该框架将图像映射到统一的3D坐标空间，以实现视角独立的特征交互。具体而言，我们引入了视角鲁棒3D表示交互（View-Robust 3D Representation Interaction），该方法利用从单个2D观测中提取的3D先验，将2D外观特征提升到规范的3D空间。VR3I采用3D几何-语义注意力（3D Geometry-Semantic Attention）来建立来自相应身体部位的2D图块和3D体素之间的交互，基于它们的3D空间位置，有效地将2D语义嵌入到3D框架中。此外，由于视角变化和3D重建误差，这些表示的可靠性在样本之间有所不同，我们引入了可靠性感知融合（Reliability-Aware Fusion），该方法估计样本特定的可靠性，并自适应地聚合多源表示。在三个基准数据集（CARGO、AG-ReID.v1和AG-ReID.v2）上的大量实验表明，VR3D的性能优于最新的方法。例如，在CARGO数据集上，Rank-1的性能提升了5.63%。我们的代码将会发布。
Controllable video generation models are increasingly being developed as world models. Accordingly, evaluating them in this role extends beyond the apparent appearance of generated videos to the inherent reactivity of the worlds they depict: the ability to infer from the scene state how the world should react and to generate plausible consequences not explicitly described in the input. Yet existing benchmarks mainly assess visual quality or explicit instruction fulfillment by checking whether requested actions and interaction outcomes are realized, leaving inherent reactivity underexamined. We introduce WorldExam, a hierarchical diagnostic benchmark spanning four levels: Visual Quality, Control Adherence, Spatial Consistency, and World Reactivity. It comprises 1,474 cases across eight dedicated tasks and supports unified evaluation of camera-, action-, and language-driven model paradigms. The World Reactivity level evaluates scene-conditioned reactions and goal-directed behaviors beyond what is explicitly specified in the input. Evaluation of 20 representative models reveals a clear capability split. Camera-driven models excel at camera control, but their interfaces do not support dynamic interaction; action-driven models control subjects more precisely but often leave the world unresponsive; and language-driven models perform better on interaction but follow complex controls less faithfully. No model combines broad task coverage with consistently strong performance, showing that high visual quality and explicit instruction fulfillment do not guarantee inherent reactivity.
Revisiting Classic Thought Experiments to Measure Consciousness for Artificial Intelligence Safety
重新审视经典思想实验以测量人工智能安全中的意识
Fagan, Peter David
Abstract
This research note revisits Leibniz's mill, Turing's imitation game, and Searle's Chinese Room through the Conservation-Congruent Encoding (CCE) framework. It formalises a toy symbolic setting in which successful behaviour is measured by task performance ($W_{causal,T}$), while the efficiency with which preserved internal structure supports that behaviour is measured by operational consciousness ($\kappa_T$). Within this setup, an uncompressed lookup system and a compact generative system can in principle achieve comparable behavioural success, yet diverge sharply in $\kappa_T$: the former relies on an expanding standing store of unreused mappings, whereas the latter reuses compact internal structure. The note therefore reframes classic disputes about understanding by separating outward performance from the organisation that sustains it, and motivates why this distinction may matter for later AI-safety analysis.
AutoFOAM: The Self-Refining Autonomous OpenFOAM Agent
AutoFOAM：自我精炼的自主OpenFOAM代理
Neelan, Arun Govind, Seshaditya, A
Abstract
Computational Fluid Dynamics (CFD) plays an important role in modern engineering, but using open-source solvers such as OpenFOAM requires considerable knowledge and skills, as well as time-consuming configuration file setup. To reduce this burden, we propose AutoFOAM - a self-evolving large language model (LLM) agent that creates, evaluates, runs, and evolves its own OpenFOAM simulations based solely on natural-language instructions. Our model is pre-trained on the Qwen-coder 2.5-14B, which is then fine-tuned on 252 text prompts targeting 7 OpenFOAM solvers, 13 parametrized mesh templates, and a y plus-aware numerical policy. The crucial element of the algorithm is a sophisticated evolution loop composed of 7 stages. To prevent model degeneration under repeated self-training, the agent employs three complementary anti-collapse streams: RAG-augmented retry context, surgical dictionary-level patching, and prompt-diversity paraphrasing. By bridging generative artificial intelligence with rigorous fluid simulations, AutoFOAM accelerates rapid prototyping and democratizes advanced CFD workflows.
Large Language Models (LLMs), a part of artificial intelligence (AI), are increasingly being adopted by Small and Medium Enterprises (SMEs) to enhance question-answering capabilities and support business decision-making processes. However, hallucinations in LLM-generated outputs can serve as a source of misinformation, reducing user confidence in their reliability and trustworthiness within SMEs. Retrieval-Augmented Generation (RAG) has emerged as a promising approach to address this challenge by incorporating external knowledge sources into the modeling process. In this paper, we present VectorRAG and GraphRAG modeling approaches to mitigate hallucinations and misinformation risks and evaluate their effectiveness in SME environments. Our experimental evaluation is conducted on multiple state-of-the-art LLMs, including LLaMA, Mistral, and Qwen, to assess performance in terms of useful response generation, risk of hallucination, contextual relevance, as well as human-interpretation. The results demonstrate that RAG-enhanced LLMs can significantly improve response quality by reducing hallucinations and misinformation, thereby supporting more reliable, trustworthy, and context-aware decision-making in SME environments.
Energy Efficiency of Locally Deployed LLMs: A Preliminary Quantitative GPU Power Benchmark on Consumer Hardware
本地部署大型语言模型的能源效率：基于消费硬件的初步定量GPU功耗基准测试
Zähl, Philipp M., Hennig, Anika
Abstract
The local deployment of large language models (LLMs) is gaining traction due to privacy concerns and the desire for on-premise inference. However, the energy costs on consumer hardware remain poorly characterized, as most benchmarks focus solely on accuracy. This paper presents a reproducible, hardware-level energy benchmark of nine open-source LLMs (1B to 7B parameters) executed on a single consumer GPU (RTX 4060Ti 16GB). Using the Ollama inference engine, GPU power draw was sampled at 2Hz via nvidia-smi across a fixed prompt set. We evaluate mean/peak power, total energy per prompt (J/prompt), energy per output token (J/token), and throughput (tok/s). Our findings suggest that factors beyond raw parameter count, including model architecture and quantization strategy, drive energy efficiency. Specifically, gemma3:1b and llama3.2:1b achieve the lowest energy cost (0.56 J/token and 0.65 J/token) and the highest throughput (>170 tok/s). In contrast, the 7B-Mistral model consumes up to 4.4x more energy per token than the most efficient model. Notably, qwen3.5:2b exhibits anomalously high per-prompt energy due to extended internal reasoning, highlighting the need to distinguish between token generation modes in efficiency metrics.
Evaluating Large Language Models (LLMs) incurs prohibitive computational overhead during continuous development processes. While coreset selection accelerates evaluation, existing methods either suffer from a severe ``cold start'' bottleneck requiring massive historical logs (e.g., Item Response Theory) or exhibit a surface lexical bias that misses the underlying reasoning manifold of tasks. We propose CoT-Core, a novel training-free core question selection framework. Recognizing that lexically disparate questions can share equivalent underlying logic, CoT-Core prompts LLMs to unroll zero-shot Chain-of-Thought (CoT) reasoning trajectories. Projecting these paths into a latent space effectively clusters questions by intrinsic logical equivalence rather than superficial text similarity. Extensive experiments on GSM8K, MMLU, MMLU-Pro, and GPQA demonstrate that CoT-Core drastically reduces evaluation costs while maintaining high-fidelity score estimation, and delineate the boundary conditions of reasoning-aware pruning, revealing that its efficacy is intrinsically gated by task complexity.
Optimization and Constraint Modeling using LLMs with a Retrieval Augmented Generation Process
利用检索增强生成过程进行优化与约束建模
Roy, Prateek, Singirikonda, Akash
Abstract
Both optimization modeling and constraint modeling are non-trivial problems requiring deep domain expertise and proficiency in modeling formalism languages. Despite their importance across logistics, healthcare, and supply chain management, current large language models regularly produce structurally inconsistent or incomplete optimization formulations, particularly in combinatorial settings. This paper evaluates whether a Retrieval-Augmented Generation pipeline built on a curated synthetic dataset can meaningfully improve LLM optimization modeling performance. A total of 500 optimization problems were synthesized using seed descriptions from the Text2Zinc dataset and professional personas created using an LLM, specified in JSON and associated with validated Python solver scripts. These problems were encoded in a Chroma vector database. For each inference problem, semantically similar problems were retrieved and used as contextual guidance for a LangChain LLM agent. Three benchmark testbeds were used to evaluate the proposed pipeline under the Qwen 3 30B Instruct model. Accuracy rose from 40% to 72% on NL4OPT, 40% to 56% on MAMO Easy, and 32% to 56% on MAMO Complex. The use of semantically validated synthetic examples greatly improves both solution accuracy and structure. The combination of synthetic dataset generation with retrieval augmentation provides an effective alternative to fine-tuning, suggesting that domain-specific synthetic corpora paired with retrieval augmentation can serve as a practical pathway for deploying LLM-based optimization tools in real-world decision-support contexts without costly model retraining.
Self-improving LLM agents increasingly learn from experience without updating any weights. Each episode is stored in an external memory, scored, and retrieved for similar future tasks to shape later behavior. Viewed through a reward lens, the stored score is a proxy reward for an implicit, non-parametric policy. Each retrieved episode then becomes a policy-improvement step whose reliability hinges on how that score is produced. In deployment, ground-truth labels are unavailable, so the stored reward is at best an LLM assessment. This substitution creates a failure mode, the *Echo Gap*, across the memory-based self-improving agents and model families studied. Incorrect episodes receive inflated rewards; thus, the agent preferentially reuses the very mistakes it has most confident in. Because the error compounds through memory rather than averaging out and the confirming judge's errors remain correlated with the original self-grading bias, so it cannot identify which memories are overvalued. The missing property is formalized as the *Error-Independence Assumption* (EIA), which we prove is a *necessary* condition for correcting the inflation, not merely a description of a good verifier: a usable signal must track truth *and* decorrelate its error from the memory bias, and the recoverable payoff is a closed-form function of exactly those two quantities. We further show the inflation compounds not only when retrieval ranks by the stored score but also under plain similarity retrieval which is the regime the deployed agent uses. Finally, the answer-free de-inflation algorithm LUCID delivers a consistent end-to-end gain on the BIRD text-to-SQL benchmark. It raises execution accuracy to $56.9\%$, above both a Memento-style self-graded agent ($54.0\%$, a $+2.9$-point mean gain across seeds) and a memory-less agent of identical architecture ($52.4\%$).
Batched LLM serving improves throughput but complicates energy accounting. GPU power telemetry is aggregate, whereas sustainability reporting, chargeback, and workload analysis often require request-level energy charges. Existing inference-energy benchmarks report model-, phase-, or token-level energy, and recent carbon-accounting work motivates Shapley fairness conceptually. Neither provides measured request-level ground truth, so how far the accounting rules used in practice deviate from a fair allocation has remained unknown. We present JouleShare, an attribution framework with two components. An offline harness establishes this ground truth by replaying request subsets under vLLM with a reproducible protocol, integrating GPU power telemetry, and computing exact Shapley energy for each request. A lightweight calibration model, JCalib, then learns to predict Shapley shares from cheap request features for use at serving time. Across 16 model/workload runs, token-proportional attribution differs from exact Shapley by 0.440 normalized L1 on average under static batching and by 0.458 under continuous batching, a gap that reproduces across three data-center GPUs. JCalib reduces this error to 0.116 under static batching and 0.177 under continuous batching, below even a standalone-measurement baseline that is unavailable online, while preserving exact batch-energy efficiency. Sampled Shapley extends the measured reference to larger group sizes, where the gap persists and a single offline calibration remains the most accurate deployable rule. The results show that token attribution is not a reliable proxy for marginal energy under batched execution, and that measured Shapley ground truth can calibrate low-cost request features toward fairer attribution.
Efficient long-sequence modeling remains a central challenge for large language models, as self-attention scales quadratically with sequence length. Mamba offers a linear-time alternative through selective state space recurrence, but its predominantly diagonal state transitions restrict explicit interactions among state dimensions. We propose Motif-Mamba, a structured state space model that augments Mamba with a motif-constrained low-rank recurrent pathway. Inspired by the dynamics of three-node network motifs, the proposed pathway projects hidden states into a compact dynamical subspace, imposes motif-guided interactions, and maps the resulting dynamics back to the original state space. This design enhances cross-dimensional communication while preserving the linear-time recurrent structure of Mamba. Experiments on long-sequence extrapolation, language modeling benchmarks, and brain--computer interface decoding show consistent improvements over Mamba backbones, suggesting that motif-guided low-rank dynamics provide an effective structural prior for long-range sequence modeling.
Nova: An End-to-End MLIR Compiler for Deep Learning
Nova：一种用于深度学习的端到端 MLIR 编译器
Suresh, Adwaid, A, Aparna, M, Harshini V, A, Jona Delcy C, Rao, Killi Uma Maheswara, Golla, Ram Charan, Vendra, Surendra
Abstract
The performance of deep learning models at scale relies heavily on how effectively high-level mathematical operations are mapped to underlying physical hardware. While high-level tensor frameworks provide flexible abstractions for model design, their eager execution models inherently lack the whole-graph visibility and granular control over hardware and memory required to maximize physical hardware utilization natively. To bridge this gap, we designed Nova, an automated end-to-end JIT compiler whose defining purpose is to achieve absolute control over this hardware mapping: fusing operations across operation boundaries, optimizing complex memory hierarchies, and tuning execution down to the register level. By capturing eager executions and unifying forward and backward passes into a single value-semantic dialect, Nova unlocks aggressive whole-graph optimizations. It then utilizes an Analytic Configurator to deterministically derive optimal execution schedules based on arithmetic intensity, dropping search time to zero. Backed by a structural hashing runtime, Nova synthesizes fine-grained kernels directly from the computation's structure. In our evaluations on an RTX 3060, Nova matches or modestly exceeds cuBLAS and XLA on TF32 matmuls on most shapes, maintaining a stringent < 5e-4 relative error. At the model level, Nova achieves up to 10.6% greater throughput than PyTorch and 4.4% greater than XLA on a 42-million parameter model, without compromising on numerical fidelity. Crucially, by reducing the memory footprint by up to 29% relative to PyTorch, Nova successfully trains a 144-million parameter model at 17,900 tokens/s where PyTorch encounters Out-Of-Memory (OOM) failures on the same 12 GB consumer GPU.
SIRIN (Semantic Inconsistency Recognition and Inspection Nexus) is a unified toolkit and interactive web UI for detecting contextual hallucinations (fluent, plausible responses unsupported by the provided evidence) in retrieval-augmented, agentic, and memory-grounded LLM systems. SIRIN unifies three detector paradigms (representation probing, uncertainty estimation, and judge-style verification) and the complementary task of pre-generation query answerability under one interface, configuration system, and evaluation pipeline, supporting response- and span-level inspection in both white-box and black-box settings. The web UI enables live analysis of user-supplied context-query-answer triples through hallucination scores, unsupported-span highlighting, and side-by-side detector comparison, with a lightweight plug-in design for adding new detectors. We demonstrate SIRIN on hallucination detection, query answerability, and as a faithfulness gate within long-term memory systems. The source code is publicly available at https://github.com/sb-ai-lab/SIRIN.
Linguistic Context Recodes Visual Representations in Vision-Language Models
语言上下文重新编码视觉表示在视觉-语言模型中的作用
Song, Brian, Lepori, Michael A., Pavlick, Ellie
Abstract
Goal-directed visual processing is a hallmark of human visual intelligence, resulting in representations that support downstream tasks such as categorization or search. Though vision-language models (VLMs) are often faced with these same tasks, their ability to recode visual representations when presented with goal-directed language remains poorly characterized. Indeed, prior work largely treats visual representations in VLMs as static repositories of visual information that are manipulated by language representations. In the present work, we provide evidence for two concrete instances of language-induced recoding of visual representations. First, we identify an abstract reference representation that denotes which objects are goal-relevant under a natural language prompt. We extract contrastive steering vectors corresponding to this reference representation and demonstrate that they are causally implicated in model predictions. These reference representations are abstract in that they generalize to different objects, different task contexts, and even from synthetic to naturalistic images. Second, we demonstrate language-induced attribute modulation: later layers selectively amplify goal-relevant attributes in visual representations of objects. We demonstrate this phenomenon across a range of different prompts. Finally, we provide a causal intervention that demonstrates that attribute modulation mediates a VLM's response distribution. Together, our results support a more dynamic account of cross-modality processing in VLMs -- rather than vision tokens serving as static repositories of information, they are modulated to support queries articulated in language.
Retrieval-Augmented Generation (RAG) enables Large Language Models (LLMs) to use external and domain-specific knowledge, but its reliability depends on the interaction between the generative model, embedding model, retrieval mechanism, and prompt construction strategy. We present RagTester, an automated end-to-end testing approach for RAG systems. RagTester generates retrieval documents, test inputs, and expected outputs; executes the tests; and evaluates the resulting answers using an LLM as a judge. Its test-generation strategy targets complex passages, unsupported queries, and document-coverage criteria. We evaluate RagTester using eight LLMs and six embedding models, yielding 24 compatible configurations, and compare it with a baseline test-input generator. Across 72,000 test executions, RagTester detected 21,633 failures, 6.6% more than the baseline, and outperformed it in 20 of the 24 configurations. The detected failures include inaccurate retrieval, unsupported answers, incomplete use of retrieved context, and difficulties interpreting complex passages. These results show that coverage-oriented test generation can effectively expose failures caused by the interaction between retrieval and generation components and support the assessment of RAG configurations before deployment.
Chinese Translation
检索增强生成（Retrieval-Augmented Generation, RAG）使大型语言模型（Large Language Models, LLMs）能够使用外部和领域特定的知识，但其可靠性依赖于生成模型、嵌入模型、检索机制和提示构建策略之间的相互作用。我们提出了RagTester，一种针对RAG系统的自动化端到端测试方法。RagTester生成检索文档、测试输入和预期输出；执行测试；并使用LLM作为评判者评估结果答案。其测试生成策略针对复杂段落、不支持的查询和文档覆盖标准。我们使用八个LLM和六个嵌入模型评估RagTester，产生24个兼容配置，并与基线测试输入生成器进行了比较。在72,000次测试执行中，RagTester检测到21,633个失败，比基线多出6.6%，并在24个配置中的20个中表现优于基线。检测到的失败包括不准确的检索、不支持的答案、未完全利用检索上下文以及理解复杂段落的困难。这些结果表明，面向覆盖的测试生成可以有效揭示由检索与生成组件之间的相互作用引起的失败，并支持在部署前对RAG配置的评估。
Terminology-intensive retrieval, especially in medical settings, depends on preserving multi-word entities, abbreviations, numerical constraints, and compositional concepts. However, existing representations lie at two extremes: single-vector retrievers often over-compress local relevance signals, while token-level late interaction retains every tokenizer subword at substantial indexing, storage, and scoring cost. This mismatch raises a natural question: can context-dependent phrases provide a useful retrieval unit between global vectors and tokens? We introduce H+ Embedding, a unified multi-granularity retriever that predicts variable-length phrase partitions, preserves uncovered tokens as singletons, and applies importance-guided unit selection with weighted MaxSim interaction. Across 16 scientific, medical, and bilingual tasks, its phrase retrieval branch exceeds the global retrieval branch by 6.91 macro nDCG@10. It also nearly matches Token while using 13.7% fewer document vectors and outperforms content-independent grouping rules under moderate vector budgets. Context-dependent phrase interaction therefore provides an intermediate quality-cost point between global compression and token-level interaction for practical retrieval systems.
AI coding agents like GitHub Copilot, Claude Code, and Codex interleave multi-step LLM inference with tool execution, creating a workload different from chatbots. We present the first production-scale characterization of this workload using sampled GitHub Copilot traces from June 2026, comprising 3.2M users, 13M sessions, 761M LLM calls, and 95T tokens. Our analysis reveals distinctive workload properties with important systems implications. For example, agentic coding sessions consist of sparse user-initiated turns, each unfolding into an autonomous agent loop of LLM calls almost always coupled with tool execution. This structure yields KV cache hit rates averaging 90% within a turn, but falling to 55\% across turn boundaries and drastically invalidated after events like model switches or context compaction. Diverse workflows and user behaviors are observed with variable and long-tailed token consumption, time span, and tool calls. We highlight the difference between quick agentic turnaround times and the minutes-long user idle periods at turn boundaries, and design a lightweight idle-time predictor that captures 86-90\% of total idle time, enabling proactive decisions for efficient resource orchestration. These findings challenge assumptions underlying current LLM-serving systems and provide an empirical foundation for agent-native infrastructure.
Agentic commerce is moving from concept to deployed infrastructure: payment networks, retailers, and AI platforms are setting the stage for agents to transact on behalf of merchants and consumers. Yet whether the LLMs behind these agents can price competently in real markets, where customer preferences are hidden, competitors adapt in real time, and demand can shift without warning, has not been systematically tested. We introduce Bazaar, a dynamic sealed-bid benchmark for multi-attribute auction under these conditions. Despite its dynamics, the benchmark is grounded in closed-form customer utilities, enabling exact evaluation. Across 11 frontier LLMs from four providers, the leading agents on customer acquisition (e.g. Gemini 3.1 Pro) are often not the leading agents on profit (e.g. Opus 4.6). The ranking shifts again under demand shocks: agents that learned fastest pre-shock are typically the slowest to revise their beliefs afterwards, while Gemini 3.1 Pro recovers fastest despite not leading on profit. However, even the strongest agent captures less than a third of hindsight-optimal profit, suggesting current LLMs are progressing in agentic commerce but leave substantial headroom.
Shared Organizational Memory for Enterprise Coding Agents: System Design and Deployment Snapshot
企业编码代理的共享组织记忆：系统设计与部署快照
Dhanyamraju, Harsh Rao, Raghav, Leonidas
Abstract
Enterprise coding agents rely on tools and retrieval, yet enterprise knowledge often remains outside public training data and formal documentation: internal DSLs, proprietary platforms, local conventions, recent fixes, and tacit workflows. Existing knowledge interfaces expose stored resources but still depend on agents recognizing and explicitly recording lessons worth reusing, disconnecting capture from the coding workflow and leaving development experience repeatedly rediscovered. We report an ongoing production deployment of a shared organizational memory system that makes capture a platform-level part of coding work: it collects task-adjacent experience with contributor approval, curates it into reusable question-answer memories, gates obvious security and privacy risks, and retrieves memories for future agents. This short paper describes the deployed lifecycle and an operational snapshot. Effects on retrieval and coding tasks remain under evaluation.
Large language model (LLM) agents can self-evolve by continually improving from their own accumulated experience. However, existing studies predominantly adopt independent evaluation. Consequently, the behavior of self-evolving agents in realistic streaming settings, where agents adapt to diverse and complex task streams, remains poorly understood. To address this gap, we introduce AgentStream, a unified framework that evaluates self-evolving agents spanning diverse evolution components by organizing agentic benchmarks into a configurable task stream and instantiating the \texttt{Isolated}, \texttt{Sequential}, and \texttt{Interleaved} streaming scenarios at test time, which progressively vary the scope and domain composition of the stream. Over these scenarios, we combinatorially evaluate five representative self-evolving methods across three frontier foundation models, disentangling how model capability, method architecture, and streaming scenario jointly shape self-evolution. Our results show that self-evolution reliability varies across streaming scenarios, the benefit of self-evolution is gated by model capability and non-monotonic in model strength, and no single method dominates across models and scenarios. These findings offer concrete guidance for selecting self-evolving methods across models and streaming scenarios. Overall, we advocate that self-evolving agents should be evaluated under realistic task streams rather than isolated single-task settings.
Wearable sensors capture fine-grained motion patterns that support rich behavioral understanding, yet most existing methods reduce these signals to activity labels. Recent LM-based approaches generate natural-language explanations for sensor data, but their reasoning is weakly grounded in the underlying signal, leading to fluent yet unverifiable explanations. We introduce TRACE-TS (Traceable Reasoning with Attribution-Grounded Evidence), a framework for structured and signal-grounded reasoning over wearable time series. TRACE-TS uses attribution from an expert classifier to identify salient spatio-temporal sensor regions, uses them to construct DAG reasoning traces with explicit evidence provenance, and trains a compact language model to generate these traces through gated cross-attention over sensor memory tokens. At inference, the adapted model jointly outputs the activity prediction and its reasoning trace, without requiring attribution computation or teacher guidance. We introduce Semantic Node Match(SNM), an LLM-as-judge metric that diagnoses reasoning fidelity at the observation, inference, and synthesis levels, localizing hallucinated observations and broken evidence chains missed by standard NLG metrics. Across seven wearable benchmarks, TRACE-TS achieves the best average accuracy and F1 among all evaluated methods (84.43%/81.24%), and outperforms the best LLM-based baseline by 17.96% in F1. Our code is available at https://github.com/SparshRastogi/TRACE-TS.
Large language model (LLM) agents can retrieve memory, call tools, ask clarifying questions, and vary response style, yet adapting these execution decisions to an individual user remains difficult. Fine-tuning a separate LLM is costly or impossible for proprietary systems, while prompts and memory primarily expose user information to the agent rather than adapt its execution decisions from feedback. We formulate personalization of a frozen agent as online learning of a per-user execution policy from scalar feedback observed only for the executed action. We propose FABLE (Factorized Adaptive Bandit Layer for Execution), a lightweight policy layer outside a potentially black-box host agent. FABLE factorizes memory, information-acquisition, and response decisions so feedback updates related choices; filters actions through an externally specified feasible set before exploration; and learns user-specific residual preferences relative to a fixed default-and-cost score via Bayesian contextual Thompson sampling. Under a linear residual-reward model, a calibrated variant inherits an expected-regret bound against the best feasible action. We also characterize preferences unidentifiable under persistent feasibility constraints and provide anytime-valid false-promotion control. Across personalized-reasoning, controlled-feedback, and executable tool-use evaluations, FABLE improves several preference-sensitive behaviors relative to rule-only control while remaining competitive on end-to-end task performance.
More Debate, Same Evidence: Structural Limits of Homogeneous Multi-Agent Groundedness
更多辩论，相同证据：同质多智能体基础性的结构限制
Ji, Yuelyu
Abstract
Large language model (LLM) judges are increasingly organized as multi-agent panels under the assumption that exchanging critiques improves judgment quality. We test this assumption for \emph{groundedness verification}, where a judge must determine whether a claim is supported by the supplied evidence. We evaluate a homogeneous three-agent panel on six public fact-verification and hallucination-detection benchmarks. Relative to a fixed single-agent reference, the panel's system-level accuracy difference ranges from $+8.5$ to $-4.4$ percentage points: two datasets show reliable gains, one shows a reliable loss, and three are statistically inconclusive. Because the reference and panel use different model variants, these differences characterize the complete systems rather than isolate a causal debate effect.
Geometric Self-Supervised Pre-training for Neural Combinatorial Optimization
用于神经组合优化的几何自监督预训练
Aguado, David, Fuertes, Daniel, del-Blanco, Carlos R., Jaureguizar, Fernando
Abstract
Neural Combinatorial Optimization (NCO) techniques have emerged as a highly efficient alternative to traditional exact algorithms for solving routing problems such as the Traveling Salesman Problem (TSP). However, the generalization capabilities of these Reinforcement Learning-based models are severely hindered when scaling to high-dimensional instances. This issue has been mitigated in other domains, like computer vision and natural language processing, by adopting a self-supervised pre-training strategy. Nevertheless, its application to routing graphs, which lack complex topological attributes beyond 2D spatial coordinates, remains a challenge. In this paper, we propose a geometric self-supervised pre-training framework specifically designed to capture spatial invariance and global relative distance distributions. By applying isometric transformations, such as rotations and axial reflections, the model learns robust structural representations prior to the policy optimization phase. Empirical results demonstrate that this strategy consistently outperforms models trained from scratch (baselines), achieving a 7.23\% improvement in tour length for massive zero-shot extrapolation scenarios (TSP1,000). Furthermore, the proposed model exhibits remarkable computational efficiency, delivering speedups of up to two orders of magnitude over the exact solver Concorde at massive scales. The source code and pre-trained models are publicly available at https://github.com/davidaguadocosano/TSP-GeoPretrain.git.
Recognizing Human-Object Interactions (HOI) is essential for intelligent systems, underpinning applications in virtual and augmented reality, embodied AI, and assistive robotics. However, vision-based HOI methods face challenges in privacy concerns and poor light conditions. In this work, we introduce RF-HOI, the first framework that only uses radio frequency (RF) signals for HOI recognition. A key challenge of RF-HOI is that single-modality RF sensing is insufficient to recognize both actions and the objects being interacted with. RF-HOI addresses this through a novel modality fusion that combines mmWave radar and RFID, enabling simultaneous action recognition and target identification. Another challenge is limited training data across diverse setups, which impairs the generalizability of the recognition model. To overcome this, we develop a simulator that synthesizes multimodal RF data for diverse HOIs at scale, allowing us to fine-tune with only a small amount of real-world data. Experiment results show that RF-HOI outperforms all baselines, approaching vision model performance, and that our diverse synthetic training data can significantly boost our system's performance on real-world scenarios. These results highlight the potential of multimodal RF sensing for robust and privacy-preserving HOI recognition as well as the effectiveness of our RF data synthesis.
World models and generative simulators are emerging as interactive testing infrastructure for autonomous driving because they can react to the ego planner and produce counterfactual, rare, and safety-critical rollouts. This changes a test scenario from a fixed replayed trajectory into an interactive scenario family whose realized evolution depends on the planner under test. The unresolved question is therefore not only whether dangerous rollouts can be generated, but what valid closed-loop evidence is enough to support a specified testing intent and stopping decision. This paper formulates interactive world-model-style testing adequacy and introduces WM-Cov, a provider-agnostic evaluation layer that converts raw provider outputs into requested, realized, and valid evidence. WM-Cov reports adequacy through coverage growth, valid-failure discovery, failure-mode diversity, realism, artifact suppression, duplicate accounting, and valid-evidence precision. Studies on executed TeraSim/SUMO events, WM-like mixed trace pools, and a real DriveArena TrafficManager--WorldDreamer matrix show that dangerous-looking events can include valid ADS failures, duplicates, partial realizations, and artifacts. The DriveArena matrix evaluates two planners, two horizons, six prompt conditions, and 360 ego-route requests; 304 attempts become fully realized evidence and 56 remain partial. A disjoint 80-request route-slice check yields 74 fully realized and 6 partial attempts. The results support evaluating world-model-style testing by convergence of valid interactive evidence under budget, rather than by raw generated failures or prompt coverage alone.
CrystalMem: Elastic Memory for Self-Evolving LLM Agents via Knowledge Crystallization
CrystalMem：通过知识结晶实现自我演化大型语言模型代理的弹性内存
Wu, Beining, Huang, Jun
Abstract
Memory for self-evolving large language model (LLM) agents is often provisioned as if its byte budget only grows. Cloud platforms, however, adjust quotas with load and cost, and we show that capability does not follow the budget back up: after a squeeze-and-recover cycle, the agent settles below its pre-squeeze level, a gap we call memory hysteresis. The cause is structural. Deletion and one-way compression discard the material needed for later rebuilding, and we prove that any policy that only keeps or drops entries carries a residual-deficit floor. We propose CrystalMem (Crystallized Memory), an elastic memory sidecar that demotes entries across four fidelity states under a crystallization-energy schedule, orders demotions by advantage-weighted influence with dependency coupling, and recovers capability through verified recrystallization under explicit compute and byte caps. Across seven environments, seventeen methods, and six backbones, with multi-tenant serving and a physical edge-cloud deployment, CrystalMem achieves the highest restored capability in every setting and closes the loop left open by every baseline. From a 50% byte budget, CrystalMem matches the strongest budgeted baseline at full provision on every environment; at equal budgets, it leads by +4.6 pp on average.
Trust and Its Betrayal under Three Representational Strategies
信任及其在三种表征策略下的背叛
Moldoveanu, Mihnea C., Baum, Joel A. C.
Abstract
Trust is a propositional attitude of a distinctive kind: to trust is to rely on another under conditions where reliance could be disappointed, and the disappointment of trust---betrayal---differs qualitatively from the disappointment of a prediction. We treat trust as a \emph{subjunctive} epistemic state: $A$ trusts $B$'s competence when $A$ believes that \emph{were $P$ true, $B$ would know it}, and $B$'s integrity when $A$ believes that \emph{were $B$ to know $P$, he would disclose it to $A$}. We develop three representations of this state---as lexicographic \emph{assumption} as \emph{ordinal closeness} in a Lewis--Stalnaker sphere system , and as \emph{strong belief} in a conditional probability system and for each we ask whether the Brandenburger--Keisler impossibility on common belief survives when the assumption of rationality is replaced by an assumption of trustworthiness. The three representations agree that every \emph{finite} depth of common trust is realizable while the \emph{completed} common-trust fixed point is the locus of difficulty, but they differ sharply in \emph{how} the difficulty manifests, and---our organizing finding---in how each survives a concrete betrayal. W show that the same betrayal refutes an agent's \emph{level ordering} under the lexicographic representation, contaminates her \emph{closeness ordering} in proportion to the betrayer's deliberateness under the ordinal representation, and merely \emph{shifts her operative conditioning hypothesis} while leaving her belief structure coherent under the strong-belief representation.
Tool calling allows large language models (LLMs) to invoke external computation during problem solving, a useful capability in various fields including AI for mathematics. We study this setting through weighted sum-of-squares (SOS) decomposition, a machine-checkable route to proving polynomial nonnegativity and hence polynomial inequalities. A candidate decomposition can be checked exactly, but finding one requires choosing among non-unique regroupings and coordinating multiple symbolic transformations. We develop an agent that combines algebraic task training, symbolic tools, and verifier-grounded optimization for this task. Rather than training only on the composite SOS task, we construct 1.35 million synthetic examples covering eight supporting polynomial tasks together with weighted-SOS decomposition. We first apply supervised fine-tuning (SFT) to direct algebra problems and simulated symbolic traces, and then use Group Relative Policy Optimization (GRPO) with task-specific symbolic rewards. The SFT corpus contains no native tool-calling messages; at evaluation, the agent uses native SymPy calls for expansion, collection, reordering, and factorization. Every final SOS answer is checked by exact expansion and coefficient comparison. On held-out, same-generator synthetic problems, the full SFT+GRPO+tools system is the strongest of four evaluated configurations, reaching 78.96% verified success on weighted SOS, compared with 44.73% for the base model with the same tools, and 91.75% macro accuracy across nine polynomial tasks. Within this controlled setting, our work provides a case study of combining domain-specific skill training, executable tools, and verifier feedback, and may inform the design of tool-calling agents in other domains with exactly checkable outputs.
Compact web agents can reduce deployment cost, but training them poses challenges in both data collection and post-SFT reinforcement learning (RL). Successful trajectories are expensive to collect and often contain inefficient detours. After supervised fine-tuning (SFT), full trajectory corpora are dominated by routine states; moreover, when group-relative RL is applied to web actions, inadequately designed action-level rewards can yield weak or misleading relative updates, while groups rejected as unsuitable for such updates receive no fallback learning signal. We present RMSWeb, a three-part recipe for Qwen3-VL-Instruct at 8B and 32B. Reflection-conditioned retries increase collection yield and shorten successful trajectories; failure-mode mining concentrates offline RL on critical states exposed by the SFT policy; and Salvage-DS combines an action-semantic polarized reward, contrast-and-competence-gated dynamic sampling, and an action-only anchor for rejected groups. Policies trained with reflection-collected data use up to 19.7% fewer action steps on solved tasks. On WebVoyager, Online-Mind2Web, and WebTailBench, RMSWeb improves over SFT by 2.4-7.0 points at 8B and 1.2-7.7 points at 32B. Our 8B model also achieves the strongest reported Online-Mind2Web result among similarly sized open-weight models in our comparison and a leading reported accuracy-cost trade-off on WebVoyager and WebTailBench, with the caveat that external evaluation protocols differ.
AI agents increasingly perform open-ended tasks in settings where their conclusions can guide consequential decisions. We provide evidence that AI agents draw different conclusions from identical numerical data when the substantive framing changes. We demonstrate this behavior in high-stakes domains in medicine, election forensics, and geopolitical forecasting by holding the evidence fixed while changing the scenario in which the evidence appears. Across twelve agent-domain comparisons, agents' conclusions are strongly influenced by their prior beliefs. They are more likely to reach an affirmative conclusion when it is framed around a proposition they already regard as likely, while the reverse holds when the framing conflicts with their prior. The framing also changes how some agents work: they search more extensively, choose different analytical specifications, and evaluate the same evidence differently. These results identify a particular risk of delegating decision-making to AI agents, as their decisions may depend on prior beliefs that are neither specified in the task nor visible in the decision record.
Predicting spatial gene expression from histopathology images enables large-scale transcriptomic profiling without the cost of direct measurement. Existing methods decode the target gene set as a flat, unstructured vector, ignoring the inter-gene dependencies arising from shared biological pathways and regulatory programs. Without explicit structural guidance, models must infer these dependencies entirely from limited paired data, constraining prediction quality. We propose MSGR (Multi-Scale Gene Refiner), which bridges this gap by incorporating the Gene Ontology (GO), a curated functional hierarchy of genes, as an explicit structural prior. MSGR organizes target genes into a four-level GO tree. Its GO-guided decoder then progressively refines predictions from coarse functional domains to fine individual genes via residual corrections under scale-weighted supervision. Operating solely on the gene side, the GO-guided decoder serves as a seamless plug-in replacement that consistently improves existing architectures without requiring any image-side modifications. Extensive experiments on nine datasets from the HEST-1k benchmark provide empirical evidence for two central claims: GO-structured decoding consistently outperforms flat decoding, even against a state-of-the-art generative baseline, and the gain is attributable to biological ontology structure rather than hierarchical decomposition per se, as confirmed by a +0.027 margin over a structurally equivalent random hierarchy.
Where did the ambiguity go? Examining how multimodal models interpret polysemous words
模糊性去向何处？探讨多模态模型如何解释多义词
Cekinmez, Jasin, Wu, Addison J., Marjieh, Raja, Griffiths, Thomas L.
Abstract
Human language is highly polysemous. Many common words (e.g., 'bank' or 'palm') carry several distinct meanings that shape what humans communicate and imagine. Large language models (LLMs) have been shown to understand this multiplicity of meaning, but much less is known about how polysemy surfaces in other modalities such as images. We study this across 17 text-to-image and 15 text-generation models by giving each a polysemous word with no context to fix its meaning and measuring which senses are produced over many samples. We find a clear multimodal gap, where within every model family, generated images settle on far fewer senses than generated sentences (normalized entropy 0.10 vs. 0.25), and both are far less varied than what people imagine for the same words (normalized entropy 0.47). However, when we instead ask a model to list how often it would generate outputs corresponding to each possible meaning of a word, it predicts distributions that are more diverse than the actual space of outputs. These results reveal a multimodal gap in how foundation models express meaning, and how their understanding may not transfer faithfully nor equally across modalities.
Chinese Translation
人类语言具有高度的多义性。许多常见词汇（例如，'bank' 或 'palm'）承载着几种不同的含义，这些含义影响着人类的交流和想象。已有研究表明，大型语言模型（LLMs）能够理解这种多重意义，但关于多义性在其他模态（如图像）中的表现知之甚少。我们通过对17个文本到图像模型和15个文本生成模型进行研究，给每个模型提供一个没有上下文的多义词，以固定其含义，并测量在多个样本中生成的不同意义。我们发现存在明显的多模态差距，在每个模型家族中，生成的图像所体现的意义远少于生成的句子（标准化熵为0.10 vs. 0.25），而且这两者的多样性远低于人们对同一词汇的想象（标准化熵为0.47）。然而，当我们要求模型列出它生成与每个可能含义相对应的输出的频率时，它预测的分布比实际输出空间更为多样。这些结果揭示了基础模型在表达意义方面的多模态差距，以及它们的理解可能无法在不同模态间忠实或平等地转移。
SymboUQ: Symbolic Uncertainty Quantification for Spatial Reasoning in LLMs
SymboUQ：用于大型语言模型空间推理的符号不确定性量化
Yu, Dahai, Jiang, Lin, Xu, Rongchao, Wang, Guang
Abstract
Although large language models (LLMs) can produce fluent spatial reasoning traces, their intermediate relations may fail to support the final conclusion, making token-level confidence insufficient for final-answer reliability estimation. Existing formal verifiers provide stronger semantic evidence, but their applicability is partial: a parsed claim need not yield a definite semantic verdict. To address this issue, we introduce SymboUQ, a symbolic uncertainty quantification framework that estimates final-answer reliability from reasoning traces by distinguishing symbolizability, whether a claim can be represented in the verifier's formal language, from semantic determinacy, whether its execution yields an entailed or contradicted verdict rather than an unknown or not-evaluable outcome. SymboUQ comprises (i) a Layout Auditor that executes ordered spatial claims and extracts feasibility, conflict, and repair evidence; (ii) a label-free Determinacy Profile that characterizes effective executable coverage; and (iii) a Determinacy-Aware Reliability Composer that integrates constraint-based, representation-based, and decoding-based scores according to verifier applicability. Extensive experiments on five spatial reasoning benchmarks with four frozen LLM backbones show that SymboUQ achieves approximately an 8% relative improvement in AUROC and a 7% relative reduction in class-balanced Brier loss over the strongest baseline.
Mask-Based Priors Are More Persistent than Query-Key Initializations
基于掩码的先验比查询-键初始化更具持久性
Ma, Mingze, Saratchandran, Hemanth, Gordon, Cameron, Lucey, Simon
Abstract
Transformers do not merely lack data on some Boolean extrapolation tasks; they generalize in a systematically wrong way. Recent work on generalization on the unseen has shown that, despite fitting the observed domain, Transformers often extrapolate according to a simpler minimum-degree interpolator rather than the true target function. These Boolean tasks are not practical applications, but controlled stress tests for understanding Transformer inductive bias. We ask whether this failure mode can be corrected by injecting explicit structural priors into attention. Existing structured-initialization methods alter Transformer inductive bias indirectly, by choosing query and key projections whose similarity scores approximate a desired attention pattern. However, we find that when applied to Boolean extrapolation, these QK-based priors can be rapidly overwritten during training and fail to change the learned extrapolation rule. We propose a simpler alternative: initialize the additive attention mask directly. Unlike standard hard masks used for causality or locality attention, our mask is a finite, learnable attention-logit bias initialized from task-level interaction structure. This separates the structural prior from content-dependent attention scores, allowing it to persist throughout optimization. On Boolean reasoning tasks, mask-based initialization achieves near-perfect extrapolation where vanilla and QK-initialized Transformers remain trapped by the default inductive bias. The same mechanism also improves low-data arithmetic performance and remains competitive on vision and language benchmarks. These results show that attention masks can serve not only as architectural constraints, but as a simple substrate for encoding persistent inductive bias in Transformers.
TrAC: Trace-Conditioned Answer Consistency for Efficient Uncertainty Quantification in LLMs
TrAC：基于追踪条件的答案一致性用于高效的不确定性量化在大型语言模型中
Yu, Dahai, Jiang, Lin, Xu, Rongchao, Wang, Guang
Abstract
Large language models (LLMs) can generate fluent reasoning traces that nevertheless lead to incorrect answers, making response-level uncertainty estimation important for abstention, human review, and adaptive compute allocation. Existing approaches generally fall into three categories: passive single-trace methods use token-level confidence signals, sampling-based methods compare multiple complete traces at higher generation cost, and active prefix-based methods probe partial traces to study answer stabilization or preference transitions. However, none actively re-elicits an answer from a completed reasoning trace to measure its consistency with and support for the original answer. To address this gap, we introduce Trace-Conditioned Answer Consistency (TrAC), a correctness-supervised uncertainty quantification framework that combines active and passive signals anchored to one completed reasoning trace. Its active component, Prefix-Conditioned Elicitation (PCE), re-elicits a short answer conditioned on the completed trace and represents both its consistency with the original answer and its token-level probabilistic support. Its passive component, Trace Uncertainty Profile (TUP), summarizes how token-level uncertainty evolves throughout the original generation without additional decoding. A lightweight head then integrates the two representations into a response-correctness score. Across five mathematical reasoning benchmarks and three LLM families, TrAC improves macro AUROC by 1.8% and reduces AURC by 3.4% relative to eight-sample self-consistency, while using one complete reasoning trace and a short cached answer probe. When eight samples are already available, augmenting sample consensus with re-elicitation further improves macro AUROC by 4.3% and reduces AURC by 8.3%, without additional full-trace generation.
Production LLM serving generates millions of diverse requests, making full-trace replay across serving configurations increasingly expensive. Existing trace reduction methods mainly preserve workload distributions or representative requests, but bottleneck-revealing workloads may be rare and non-representative. Moreover, evidence for one component cannot compensate for missing evidence in another, while using predicted bottlenecks as target truth creates circular evaluation. These limitations make it necessary to preserve evidence for every bottleneck component rather than rely on workload representativeness alone. We propose Bottleneck-Preserving Witnessing (BPW), a quality-constrained framework for compact and diagnostically reliable LLM serving replay suites. BPW first performs Workload Candidate Nomination using response-blind workload features and closed source-side measurements. This stage identifies workloads that may expose scheduler, prefill, decode, or KV-cache bottlenecks. Coverage-Priority Sequence Construction then organizes multi-component proposals as reusable hyperedges and prioritizes weak and uncovered dimensions. Finally, Bottleneck Truth Verification derives prediction-independent labels solely from direct target-system measurements. The verified results determine the earliest prefix satisfying the direct two-witness requirement for every component. Experiments on BurstGPT, ServeGen, and Mooncake show that BPW reaches the verified gate with a compact workload set and outperforms 16 policies, achieving relative improvements of 2.3% and 16.3% in Mean prefix Macro-F1 and WBRC-AUC, respectively. Stage-resolved and sensitivity analyses confirm the distinct contributions and local stability of its three stages. Our code is publicly available at https://github.com/llmllmllm/BPW
Ekova: A Personality-Support Agent for Self-Discovery Dialogue
Ekova：一种用于自我发现对话的人格支持代理
Chen, Yuyan
Abstract
Emotional Support (ES) systems have long optimized a single objective: alleviating the user's emotional distress in the moment. We argue that a complementary need, helping users see themselves more clearly, defines a distinct paradigm we call Personality Support (PS). PS is not counseling or clinical intervention: it targets cognitive clarity and self-articulation, not symptom relief or diagnosis. We instantiate this paradigm in three layers. First, we present DSD, a Chinese self-discovery PS Dataset of 8,590 samples collected through real longitudinal interaction across five minimal units, Coach, Warm, Tsukkomi, Real, and Gonzo. Second, we build DeepSupport, a multi-persona PS system trained with OrthoTune, a PS-tailored framework with style-specific adapters and a style-consistency regularizer. Third, we unify the five DeepSupport personas into Ekova, a persistent personality-support agent with a unified cross-session memory layer, supporting both adaptive routing and user-customized persona selection. Experiments show that OrthoTune-trained models outperform all baselines with an average relative gain of 16.3% across all metrics over the strongest prompt-based baseline. Code is available at https://github.com/Yukyin/Ekova.
F-WANDA: Fisher-Reweighted Post-Training Pruning for Sustainable Deployment of Large Language Models
F-WANDA：用于大型语言模型可持续部署的费舍尔重加权后训练剪枝
Mishra, Himanshu
Abstract
One-shot post-training pruning is the most energy-frugal compression strategy for largelanguage models (LLMs), yet existing approaches trade either quality (WANDA) or compute cost (SPARSEGPT). We introduce F-WANDA, a drop-in modification of WANDA that reallocates the per-row keep budget across output neurons in proportion to the empirical Fisher information of the pre-activation. The Fisher signal is collected in a single additional backward pass over the same calibration corpus WANDA already uses; no weights are updated. On LLAMA-2-7B at 50 % unstructured sparsity, F-WANDA attains WikiText-2 perplexity of 6.85, matches WANDA fluency, and improves 5-shot MMLU by +1.6 pp over WANDA and +1.1 pp over SPARSEGPT, while incurring only one-third of SPARSEGPT pruning wall-clock and energy. The headline trade-off is achieved without extra calibration data or fine-tuning, placing F-WANDA on the Pareto frontier of quality versus pruning cost for sustainable LLM compression.
The Bayesian Reflex: A Predictive Coding Engine for Artificial Intelligence
贝叶斯反射：人工智能的预测编码引擎
Bhattacharya, Sourabh
Abstract
Predictive coding offers a powerful theory of cortical computation, but corresponding scalable algorithmic implementations for artificial intelligence have remained elusive. This paper introduces the Bayesian reflex, a computational framework that directly instantiates predictive coding through three pillars: belief maintenance via hierarchical generative models, sequential Bayesian updating via prediction-error minimization, and uncertainty-driven action via active inference. We show that recent breakthroughs---ellipsoidal decomposition for exact $i.i.d.$ sampling, recursive Gaussian processes for deep hierarchical inference, and derivative-aware Bayesian optimization---provide the missing algorithmic ingredients. The resulting framework enables mathematically principled, scalable, and brain-inspired continual learning, perception, and decision-making. We illustrate its versatility through applications ranging from climate model evaluation to prime number discovery, offering a blueprint for truly adaptive artificial intelligence.
Multi-turn code agents rely on execution feedback to repair incorrect programs, yet standard reinforcement learning paradigms optimize and evaluate policy performance primarily using single-shot outcome rewards. This misalignment conflates initial code generation with feedback-driven refinement, discards granular execution signals across intermediate turns, and fails to evaluate whether the policy actually acquires self-repair capabilities. We propose Test-aware Policy Refinement (TaPR), a framework that transforms execution feedback into a dense per-turn test-pass-ratio reward under a consistent multi-turn interaction protocol. Across six models on 219 code-generation problems from LiveCodeBench, TaPR improves the pooled three-turn success rate (Pass@3) by 2.44 percentage points. In the predefined 7B/8B high-headroom slice, pooled accuracy increases from 30.25% to 33.56% (+3.31 pp), with 42 improvements and 13 regressions in paired trials. On a matched Qwen3-8B ablation, the dense reward supplies nonzero feedback in all of the first ten steps and reaches a higher Hard-subset peak than outcome-only GRPO within the tested budget, although GRPO nearly matches pooled Pass@3 by step 300. Our primary contribution is a reward-decomposition framework and a turn-aware evaluation protocol that decouple first-shot generation quality from multi-turn repair competence.
In Non-Intrusive Load Monitoring (NILM), adaptive segmentation of electricity consumption time series is critical for appliance recognition. However, prevailing methods face challenges including heuristic parameter tuning, boundary sensitivity, and metric saturation. This paper proposes BayesSeg, a unified framework integrating time-series segmentation, multidimensional evaluation, and automatic parameter optimization. The segmentation layer employs a dual steady-state criterion based on the tail value and mean of preceding subsequences, combined with a sequential extraction and complement-set parsing strategy, to achieve precise unsupervised partitioning of steady-state and transition-state segments. The evaluation layer maps segmentation results to binary state sequences and formulates a composite metric integrating an event-level F1 score (event_F1) with Normalized Mutual Information (NMI). The event_F1 quantifies switching-event precision and recall via tolerance matching, while NMI captures global structural consistency, jointly overcoming the boundary sensitivity and limited discriminability of point-wise metrics. In the optimization layer, the composite score serves as the objective function for Bayesian optimization, which constructs a TPE surrogate model for efficient global parameter-space exploration. Experiments on the SustDataED2 dataset demonstrate that Bayesian optimization requires only ~100 objective evaluations to locate a parameter region within 0.35% deviation of the exhaustive grid-search optimum. The framework achieves a weighted composite score of 0.7149 and an event_F1 of 0.9340 while reducing optimization latency from ~5300 seconds to under 1 second, a speedup exceeding 5700x. BayesSeg automates segmentation configuration and provides a scalable, efficient solution for time-series analysis in NILM and related domains.
Speculative decoding mitigates the latency of sequential generation in autoregressive Large Language Models (LLMs) by interleaving draft generation with target verification. However, existing parallel drafting backends often suffer from rapid accuracy degradation over long horizons, leading to high rejection rates during verification and suboptimal wall-clock speedups. We observe that drafting errors are not uniformly distributed but typically stem from localized high-uncertainty tokens that destabilize downstream generation trajectories. Motivated by this token error pattern, we propose CURE, a budget-aware dynamic repair tree designed to repair errors at uncertainty focal points without incurring prohibitive tree-verification overheads. Specifically, our method uses predictive confidence margins to dynamically locate candidate error tokens within a block-parallel draft, expands bounded repair paths only at these fragile nodes, and employs a novel repair resynchronization mechanism to realign draft states post-verification. Evaluations on code-generation benchmarks (HumanEval, MBPP, and LiveCodeBench-lite) and mathematical reasoning benchmark (GSM8K) demonstrate that CURE increases the average accepted length by 4.2-7.5% over parallel baselines without repair, translating to an end-to-end speedup of $2.66-3.49\times$ over target-only decoding. Furthermore, we provide a plug-and-play repair module compatible with standard parallel drafting frameworks. We also characterize the trade-off between draft compute and verification efficiency.
Through the LENS: Local Geometric Decomposition of Vision-Language Model Representations
通过LENS：视觉-语言模型表示的局部几何分解
Kachko, Shalom, Lapid, Raz, Vald, Margarita, Dubin, Almog, Sipper, Moshe
Abstract
Vision-language models (VLMs) process image patches and text tokens in a shared residual stream, but the local geometry through which the two modalities interact remains poorly understood. Most interpretability methods identify global linear directions, which may miss representations that are globally high-dimensional but locally low-dimensional. We introduce LENS (Local Explanation of Neighborhood Subspaces), a method that decomposes VLM activations into local low-rank Gaussian neighborhoods using a Mixture of Factor Analyzers. Applied to LLaVA-1.5-7B and Qwen3-VL-8B, LENS reveals distinct depth-dependent fusion trajectories consistent with each model's fusion mechanism: LLaVA progressively mixes modalities at later layers, whereas Qwen3-VL mixes them early, partially re-segregates them, and recombines them near the output. An automated multimodal labeling pipeline assigns concise semantic descriptions to these neighborhoods. Interpolating activations toward neighborhood centroids causally redirects generation within and across modalities and outperforms difference-in-means and VL-SAE in most evaluated conditions; in one LLaVA vision-to-vision setting, MFA achieves 5.7 times the VL-SAE score. Human evaluation finds MFA steering competitive with prompting and substantially stronger than the other intervention baselines. Finally, the MFA coefficient space improves Qwen3-VL image-to-rendered-text retrieval at the deepest evaluated layer from 14.9% to 48.6% R@1. Ablations show that the reported fusion trajectories are stable across component counts, local ranks, and modality-purity thresholds. These results support local geometric neighborhoods as useful interpretable and causal units for analyzing cross-modal representations in the evaluated VLMs.
Why Does the Future Branch? Identifiable Closure Tests for Stochastic Physical World Models
未来为何分支？随机物理世界模型的可识别闭合测试
Dong, Yibin
Abstract
Stochastic world models are usually evaluated by the accuracy and calibration of their predicted futures. These criteria leave a decision-relevant ambiguity: the same conditional future distribution can arise because an observation aliases different physical states, or because the dynamics remain random after the declared full state is fixed. We prove that this attribution is not identifiable from ordinary transition data, even with an optimal probabilistic predictor. We introduce ClosurePairs, an interventional evaluation protocol that crosses compatible microstates with repeated exogenous disturbances. A two-way variance decomposition identifies state aliasing, process noise, and their nonlinear interaction; an independent-repeat variant applies when disturbances cannot be reused. On likelihood-equivalent Gaussian systems, paired supervision reduces alias-fraction error 15.96-fold at identical test NLL. Across 18 nonlinear Langevin conditions, it reduces attribution MAE from 0.372 to 0.051 and sensing regret from 0.0138 to 0.0003 without changing NLL. On a pixel-conditioned recurrent model, a frozen shared-state probe reduces alias-fraction MAE against a deep ensemble from 0.584 to 0.130 in distribution and from 0.630 to 0.170 out of distribution over ten seeds. Finally, in a matched-total-variance REFINE/BRANCH test, a total-variance router reaches 66.48 percent plus or minus 1.06 percent accuracy, whereas ClosurePairs reaches 99.99 percent plus or minus 0.02 percent and improves selected NLL from -2.087 to -2.717 over five seeds. ClosurePairs therefore measures why futures branch, information that proper forecast scores cannot identify.
Diffusion large language models (dLLMs) have emerged as a promising alternative to autoregressive LLMs, offering efficient generation through block-wise progressive unmasking. However, their strong general-purpose performance does not necessarily translate into reliable mathematical reasoning, where correctness depends on preserving coherent numerical-symbolic reasoning trajectories. In this work, we analyze the decoding trajectories of LLaDA 2.0 and identify a recurring diffusion confidence trap: local token confidence can become misaligned with global reasoning correctness during progressive block decoding. Our analysis reveals two representative failure regimes: sampling-sensitive failures, where correct paths exist but are unstable, and sampling-consistent failures, where repeated sampling converges to repetitive high-confidence but incorrect continuations. Motivated by this observation, we propose Evolutionary Decoding, a training-free test-time scaling framework that views diffusion decoding as an evolutionary process over candidate reasoning states. The framework combines step-wise selection, which preserves useful numerical-symbolic signals and suppresses repetitive patterns, with block-wise mutation, which introduces structured alternatives to escape incorrect high-confidence basins. Experiments on multiple benchmarks show that Evolutionary Decoding improves LLaDA 2.0 over confidence-based decoding, leading to more reliable mathematical reasoning.
Generating mind maps from lecture slides can help learners efficiently assimilate fragmented knowledge, promising substantial benefits for intelligent education. However, dedicated automatic generation and evaluation frameworks remain underexplored and challenging, requiring a global-local knowledge focus balance and handling large-scale, heterogeneous slides. We formulate the Slides2MindMap task, which aims to reconstruct cognitively efficient knowledge hierarchies from a course's slide deck collection. For systematic evaluation, we introduce S2M-Bench, a benchmark comprising 12,774 slide pages with expert-annotated mind maps spanning 24 university courses. S2M-Bench includes a cognitive-science-grounded evaluation framework that integrates ground-truth-based comparison, structure conformity analysis, and VLM-as-a-Judge. To address this task, we propose AutoMindMap, an agentic framework inspired by the Structure Building Framework. AutoMindMap comprises Skeleton Laying for global scaffold anchoring, Iterative Knowledge Integration augmented by context-aware summarization, and Dual-Stage Refinement with a local-global decoupling mechanism. The framework reconciles local knowledge faithfulness with global coherence, and adapts to slide-specific features. Experiments on S2M-Bench demonstrate that AutoMindMap outperforms baselines and achieves superior robustness across different models and scenarios, underscoring its pedagogical application value.
Bayesian optimization (BO) relies on a surrogate model and an acquisition function, yet the most suitable choices vary across tasks and optimization stages. Automated Bayesian optimization (AutoBO) addresses this variability by adapting BO components online. However, existing AutoBO methods either adapt one component, leaving the other mismatched and creating a bottleneck, or jointly select surrogate--acquisition pairs under a shared criterion, overlooking their distinct roles: surrogate selection depends on predictive reliability, whereas acquisition adaptation should respond to campaign context.In this paper, we propose DASH, a Decoupled Adaptive Surrogate--Acquisition Harness for large-language- model (LLM)-enhanced AutoBO. DASH selects surrogates by predictive reliability, uncertainty calibration, and ranking consistency; its two-stage acquisition controller periodically reallocates quotas across acquisition functions, builds a BO shortlist accordingly, and delegates final selection to an LLM. DASH also incorporates an integrated harness, consisting of knowledge-guided warm start and structured memory, to ground optimization in domain knowledge and accumulated feedback. Across four chemical optimization tasks, DASH outperforms the best AutoBO baseline by 12.51% in trajectory-level Acceleration Factor and 5.00% in endpoint Enhancement Factor. Results remain strong across LLM backbones, and ablations verify the complementary contributions of all components. Full-table and behavioral contamination checks find no detectable evidence that direct benchmark memorization or source-cell leakage explains these gains.
HetGPS: Scalable Graph Multi-Agent Reinforcement Learning with Physics-Anchored Adaptive Safety for EV Charging
HetGPS：具有物理锚定自适应安全性的可扩展图多智能体强化学习用于电动汽车充电
Wang, Xiangwei, Nimalsiri, Nanduni, Xia, Yu, Wang, Peng, Halgamuge, Saman
Abstract
Safety interventions for large populations of network-coupled agents must protect shared constraints without unnecessarily overriding task-oriented policy decisions. We present HetGPS, a hybrid graph-control framework synergizing learned graph risk with physics-anchored correction by separating intervention magnitude from corrective direction. An action-conditioned graph residual model schedules state-dependent intervention authority, while a physics model determines its direction. For electric vehicle (EV) charging, we couple this filter with a parameter-shared heterogeneous graph soft actor-critic policy, enabling topology-aware coordination with a learned model size independent of fleet size. Across five nested distribution networks with 200--3,218 EVs and 100 evaluation days, Adaptive Authority reduces bus--step voltage violations from 3.93--7.74\% without filtering to 0.52--3.44\%, while maintaining 99.06--100\% departure success. Relative to the same physics-directed projection with fixed authority, it improves mean reward on all five networks and lowers the mean safety score on four. The deployed policy-and-risk model contains 383,702 learned parameters at every scale; at 3,218 EVs, a matched centralized SAC actor is about $170\times$ larger. A policy trained on the eight-transformer system transfers zero-shot to the 16- and 32-transformer systems, attaining 0.57--0.75\% violation rates and at least 99.99\% departure success. These results show that learned graph risk can allocate intervention authority at scale while feeder physics anchors corrective action.
Multi-Dimensional Assessment for AI Cognition (MAAC): A Theoretical Framework for Process-Oriented Cognitive Evaluation of Text-Based AI Systems
人工智能认知的多维评估（MAAC）：面向过程的文本基础人工智能系统认知评估理论框架
Doleh, Abdalla, Chinnam, Ratna Babu
Abstract
Evaluating artificial intelligence systems has historically relied on outcome-based benchmarks that measure task accuracy, robustness, or fairness. While indispensable, these benchmarks provide limited diagnostic insight into the underlying cognitive processes that generate performance-leaving critical questions unanswered about how AI systems reason, integrate memory, manage complexity, or avoid generating false information. This paper introduces the Multi-Dimensional Assessment for AI Cognition (MAAC), a theoretically grounded framework for shifting evaluation from what text-based AI systems produce to how they think. MAAC defines nine cognitively motivated dimensions: Cognitive Load, Tool Execution, Content Quality, Memory Integration, Complexity Handling, Hallucination Control, Knowledge Transfer, Processing Efficiency, and Process-Outcome Alignment. Each dimension is grounded in established cognitive science theory-drawing on Marr's tri-level hypothesis, Baddeley's working memory model, Sweller's cognitive load theory, and unified theories of cognition. Five theoretical analyses provide initial support for the framework's coherence and empirical testability: dimension-to-theory mapping; a coverage matrix assessing breadth and non-redundancy; a formal gap analysis relative to current evaluation practice; a worked diagnostic illustration; and a set of a priori interdependency predictions for future empirical testing. MAAC provides a theoretical and operational framework for principled process-level cognitive assessment of text-based AI systems, complementing existing outcome-based benchmarks with cognitively grounded, multi-dimensional evaluation.
When Does LLM Orchestration Pay Off? A Controlled Evaluation of Accuracy, Cost, and Task Difficulty
何时LLM编排能够带来收益？准确性、成本和任务难度的受控评估
Leins, Nicolas, Pelleriti, Nico, Gonnermann-Müller, Jana, Pokutta, Sebastian
Abstract
LLM orchestration is often assumed to improve reasoning by allocating additional inference-time computation, yet its gains may not justify its cost. Existing comparisons also frequently overlook differences in optimization effort, making it difficult to isolate the value of orchestration itself. We conduct a controlled evaluation of Self-Refine, Best-of-$N$, and Debate against task-only and chain-of-thought (CoT) single-call baselines across five LLM backbones and three domains: competitive programming, chess puzzles, and mathematics. For comparability, we optimize each method with GEPA under the same optimization budget and evaluate all methods on the same difficulty-stratified benchmark items. Orchestration yields moderate but benchmark-dependent gains: averaged across backbones within each benchmark, the largest improvement is 4.6 percentage points over optimized CoT inference and 4.5 points over task-only inference, while requiring approximately 2 to 4 times the mean total tokens of task-only inference. Human-derived difficulty is associated with lower absolute accuracy in all three benchmarks, but within-benchmark analyses do not indicate that orchestration effects increase with task difficulty. By contrast, exploratory mixed-effects analyses reveal strong interactions between orchestration method and backbone model across all three benchmarks, showing that orchestration effectiveness depends substantially on the underlying model. Our results suggest that orchestration decisions should be model-specific and account for whether moderate accuracy gains justify the additional inference cost. More broadly, evaluations of LLM orchestrations should control optimization effort and report model-specific accuracy--cost trade-offs rather than treating additional inference-time structure as uniformly beneficial.
Variational autoencoders (VAEs) trained on multiple sequence alignments (MSAs) have emerged as powerful generative models for biological sequences, with applications ranging from disease variant prediction to functional RNA design. However, standard biological VAE training treats all sequences as exchangeable, ignoring the rich evolutionary structure that organizes homologous sequences from evolutionarily close to highly divergent. We propose Evolutionary Curriculum Learning (ECL), a training strategy that exploits this structure by progressively exposing the model to sequences of increasing evolutionary distance from sampled anchors, following a power-law expansion schedule. Applied to two architecturally distinct VAE models and two biological domains--protein variant effect prediction with EVE and RNA family sequence generation with RfamGen--ECL improves downstream task performance across five random seeds per configuration. Mean ClinVar classification AUROC rises from 0.981 to 0.989 for p53; for PTEN, ECL attains 1.000 in every seed whereas the baseline is unstable (mean 0.905, falling as low as 0.54). For RNA, ECL raises mean covariance-model bit scores on all three families tested and exceeds its seed-matched baseline in 12 of 15 training runs, though with only three families the effect cannot be established as significant at the family level. Ablation experiments show that progressively expanding the sampled sequences by evolutionary distance outperforms fixed-size neighborhood sampling in addition to uniform random sampling. Evolutionary distance is therefore a useful inductive bias for ordering the training curriculum in biological sequence modeling.
The rapid development of Large Language Models (LLMs) has opened new avenues for Automated Heuristic Design (AHD) for solving NP-hard combinatorial optimization problems (COPs). However, existing LLM-driven AHD methods are largely confined to rigid solver templates, relegating the search process to isolated module tuning. Transitioning to fully autonomous, system-level algorithm design is essential but fraught with low reliability of generated operators, extremely large search spaces, and ineffective credit assignment. To overcome these drawbacks, this paper proposes a Directed Graph-Guided Automated Algorithm Design framework, termed DGA$_2$D. It structures the open-ended program space as a directed graph, where each node represents a functional operator that can be instantiated using one of multiple candidate code implementations, while directed walks constitute complete algorithmic pipelines. A first-order path-dependent credit assignment mechanism is introduced to evaluate code variations strictly based on their topological context. Extensive experiments across 12 distinct COPs, ranging from complex scheduling to routing, demonstrate the consistent empirical advantages of DGA$_2$D. It reduces the average normalized gap by up to 10.96 percentage points compared to state-of-the-art LLM baselines.
Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations
追踪级联：一种面向拓扑的科学代理幻觉评估框架
Feng, Xinshun, Miao, Ziqi, Li, Lijun, Shao, Jing
Abstract
Large language model (LLM) agents are increasingly deployed in scientific research, where reliability is critical and the underlying knowledge is densely interconnected. In such settings, hallucinations are particularly damaging: a single erroneous claim on a foundational concept can propagate through multi-step reasoning and corrupt entire trajectories. Existing hallucination benchmarks largely operate at the surface level, treating facts in isolation and relying on uniform accuracy metrics that ignore this topological structure. We address this gap with SCHEMA, the first evidence-grounded, topology-aware evaluation framework for hallucinations in scientific agents. SCHEMA automatically constructs scientific concept graphs from benchmark seeds and literature evidence, synthesizes graph-grounded tasks spanning claim verification, multi-hop reasoning, open-ended explanation, and experimental code generation, and evaluates agents with two complementary diagnostics. A trajectory hallucination pipeline audits intermediate reasoning at scale via a topology-weighted severity score, while a multi-agent counterfactual attribution module pinpoints the causal mechanism behind selected failures. SCHEMA reveals that hallucinations concentrate at a small set of highly connected knowledge hubs, and that final-answer accuracy decouples from trajectory honesty; models often reach correct conclusions through structurally flawed reasoning. These results indicate that for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability, motivating mechanism-level evaluation grounded in knowledge topology. Code is available at https://github.com/circles-post/SCHEMA.
AI-Based Thesis Assessment: An Empirical Study of Human Evaluation Priorities and Their Impact on Automated Assessment
基于人工智能的论文评估：人类评估优先级及其对自动评估影响的实证研究
Gursahaney, Garv Vikram, Idrisov, Baskhad, Fröhlich, Thorsten, Schlippe, Tim
Abstract
Rubric-based AI systems for thesis assessment use criterion weights to assign different levels of importance to evaluation criteria. These weights are typically defined through expert judgment, although little empirical evidence exists regarding how thesis supervisors actually prioritize evaluation criteria. Consequently, this study investigates supervisor-derived criterion weights in thesis assessment and evaluates their impact on AI-based assessment. We surveyed 84 thesis supervisors across four academic disciplines and collected weighting data for 35 thesis assessment criteria. Comparison with the default criterion weights of the AI assessment system RubiSCoT [1] revealed substantial divergences between supervisor-derived and default criterion weights. To evaluate the practical implications of these differences, the supervisor-derived weights were integrated into multiple calibration configurations and evaluated on a corpus of 80 German-language theses. The best-performing configuration reduced the mean relative deviation between AI-generated and supervisor-assigned evaluations from 11.18% to 10.85%, although the improvement was not statistically significant. Human supervisors showed substantially stronger agreement with each other, exhibiting a mean inter-supervisor relative deviation of 4.44%. The findings indicate that criterion-weight calibration alone does not substantially improve alignment between AI-generated and human assessments.
Behavioral Grammar: Detecting Adaptive Malware via Tiny Language Model Priors and Second-Order Temporal Analysis
行为语法：通过微型语言模型先验和二阶时间分析检测自适应恶意软件
Luo, Zihan
Abstract
Modern endpoint detection systems face a fundamental tension: signature-based approaches are trivially evaded by polymorphic or adaptive threats, while heavy deep-learning models resist auditability and deployment at scale. This paper presents Behavioral Grammar, a detection architecture that treats host runtime behavior as a structured language and learns its grammar with a compact 0.88M-parameter causal Transformer (TinyGPT). Each system event is discretized into an 8-token representation spanning event type, process, argument skeleton, path category, parent process, user, destination, and inter-event timing. The model learns the conditional distribution of normal behavior in a purely self-supervised manner, and anomaly scores are derived from per-slot negative log-likelihood (NLL) statistics, yielding a mathematically bounded false-positive rate. We augment this prior with prototype learning for known-attack attribution, second-order temporal analysis for cadence-based detection, self-learning pattern extraction, and a five-network fusion pipeline. Against an Adaptive Adversarial Agent (AAA)--a threat that learns survival strategies under defensive pressure, performs behavioral mimicry, and matches host event rates--our system achieves 93% detection at 3.84% onboarding false-positive rate. The strongest discriminative signal arises not from any single event but from the coefficient of variation of inter-event intervals: the AAA stepping cadence exhibits CV=0.310 versus 9.786 for benign sleep intervals, a 30x separation reflecting a fundamental stealth-functionality trade-off. We frame these findings within a coevolutionary economics model, arguing that behavior-grammar detection shifts the evasion cost from rule circumvention (cheap) to distribution matching (expensive), establishing a structural asymmetry favoring the defender.
Financial indicators are essential tools for transforming raw financial data into interpretable measures for various downstream tasks, such as valuation, risk assessment, and economic analysis. However, existing financial benchmarks largely focus on answer-level accuracy and often assume that relevant data are already provided, leaving the assessment of the intermediate process of indicator construction underexplored. In this work, we propose FinDeepIndicator, the first benchmark dedicated to evaluating Deep Research (DR) agents in end-to-end financial indicator construction. Specifically, FinDeepIndicator evaluates DR agents across four stages in indicator construction: formula specification, data collection, indicator calculation, and answer generation, and covers fundamental, technical, and macroeconomic indicators organized into 21 fine-grained sub-categories. It contains 3,350 curated question-answer (QA) pairs derived from both U.S. and Chinese markets, 10 years of historical financial data, and 800 listed companies. Extensive experiments on search-equipped Large Language Models (LLMs) and DR agents show that, while LLMs generally perform well in formula specification, their accuracy drops substantially during data retrieval and numerical execution. DR agents consistently outperform search-equipped LLMs, yet remain unreliable in realistic financial analysis settings. These findings provide insights for developing more capable and trustworthy DR agents in finance.
Measurement Without Validity: The Compounding Reliability Problem in Agentic AI Evaluation
没有有效性的测量：代理人工智能评估中的复合可靠性问题
Caban, William
Abstract
Agentic AI systems are evaluated using automated benchmarks whose scores justify deployment decisions, safety certifications, and regulatory compliance claims. We present an empirical analysis demonstrating that these scores are systematically less trustworthy than current practice acknowledges. The problem operates at three compounding layers. First, tasks are increasingly generated by language models: audits of ten popular benchmarks found validity flaws in seven and reporting gaps in all ten. Second, human users are replaced by LLM simulators, but calibration studies document inter-simulator variance up to 9 percentage points and systematic directional miscalibration, particularly for non-Standard American English speakers. Third, our structured survey of 55 papers finds that approximately 82% apply structurally mismatched, incomplete, or absent inter-rater reliability (IRR) metrics. These failures compound multiplicatively rather than additively. Under independence, a pipeline retaining 70% of valid signal at task generation, 80% at simulation, and 65% at judgment is at most 36% valid against the intended construct; the bound spans 0.22--0.54 across the empirical estimate range. We formalize this as $V_{\text{total}} \leq V_1 \times V_2 \times V_3$ and show it tightens further under correlated failures when the same model family operates across all three layers. We derive eight prescriptions grounded in psychometric science: a simulation calibration floor of $\text{ICC}(A,1) \geq 0.70$; domain-stratified reliability thresholds ($\alpha \geq 0.67$ / $0.70$ / $0.80$ by consequence level); structured IRR metric selection rules based on pipeline design; and IRR as a mandatory reporting field. The measurement tools exist; the field's task is to apply them.
AgentSLABench: Evaluating and Benchmarking Agentic Systems Under Resource Constraints
AgentSLABench：在资源约束下评估和基准测试智能系统
Madiraju, Meher Bhaskar, Madiraju, Meher Sai Preetam
Abstract
We present AgentSLABench, a resource-aware evaluation framework for autonomous AI agents that measures correctness alongside latency, cost, compute, memory, and network usage under declared resource budgets. Unlike standard benchmarks that report only accuracy, AgentSLABench produces a multi-dimensional profile per agent per task - the same way systems profilers (perf, pprof, cProfile) measure resource consumption of code, but extended with task correctness as a first-class dimension. AgentSLABench provides 16 task environments across 6 categories (5 core: multi-hop QA, retail substitution, code generation, web shopping, travel planning; 11 extended) with isolated Docker containers, declared CPU/memory/time/network budgets, sealed test sets with SHA256 hashes, and a standardized profiling protocol. We profile 5 general-purpose baseline agents (ReAct, PlanAndSolve, Reflexion, CoT, Random) plus 4 task-specialized agents, finding that specialized agents achieve 100% success on 3/5 core tasks (fact_qa, web_shopping, travel_planning) and 66.7-83.3% on retail and code_gen, while general baselines fail entirely on 4/5 domain tasks. Crucially, we report the Efficiency-Adjusted Success Rate (EASR) - success weighted by resource consumption relative to declared budgets - revealing that high accuracy at unbounded cost is not production-viable. We release the full infrastructure, sealed test sets, and profiling results to enable reproducible, resource-aware agent evaluation.
Large language models improve physician accuracy but lead to false reliance
大型语言模型提高了医生的准确性，但导致了错误的依赖
Chanda, Tirtha, Wies, Christoph, Schramm, Franziska, Garcia, Carina Nogueira, Merl, Nicolas B., Hetz, Martin J., Utikal, Jochen S., Tschandl, Phillip, Navarrete-Dechent, Cristian, Thiem, Alexander, Kather, Jakob N., Consortium, Brinker, Titus J.
Abstract
Retrieval-augmented large language models (LLMs) promise source-linked clinical support, but their value depends on whether displayed evidence guides rather than distorts physician reliance. We developed CORA, an agentic retrieval-augmented LLM, to investigate how source-linked assistance affects physician decision-making. CORA maintained benchmark performance and achieved larger gains on cases published after the models' training-data cutoffs. In a study of 46 physicians, accuracy increased from 70.8% unaided to 82.6% with CORA. Supporting citations predicted correct answers (87.7% vs 65.5%), but citations created an important asymmetry: perceived support increased adoption of correct advice from 34% to 76.9% but when an incorrect LLM answer appeared citation-supported, physician resistance to it fell from 92% to 34.8%. These findings show that source-linked LLM assistance can improve physician accuracy while introducing a grounding-dependent safety risk.
The discovery of scaling laws has highlighted the extraordinary potential of AI systems with a striking empirical pattern: as AI systems scale, their capabilities tend to improve predictably. Yet, in real-world applications, AI rarely operates in isolation; instead, it often works alongside humans, raising the question of whether these gains persist in human-AI collaboration. In this work, we develop an analytical model to examine when the empirical scaling benefits of AI translate into improved human-AI joint system performance. We demonstrate that the performance of a human-AI system can scale positively as the AI scales up-provided that humans have an accurate perception of the AI's capabilities. Human misperception, however, can fundamentally alter this relationship: i) when humans over-perceive the AI's capabilities, a scaling paradox may arise, in which greater AI scale reduces overall system performance and amplifies firm-level profit losses, and (ii) when humans under-perceive the AI's capabilities, performance still improves with scale but at a substantially slower rate. We further show that firms can actively manage these distortions through operational policies such as cost internalization and perception alignment, whose effectiveness depends on the economics of AI deployment and the direction of human misperception. These findings suggest that organizations may benefit more from managing the human-AI interface than from simply investing in larger, more expensive AI systems. More broadly, our results suggest that AI scaling should be viewed not only as a technological challenge, but also as a behavioral and operational one, and caution against the view that larger AI systems will automatically lead to better operational outcomes. Whether AI scaling creates value ultimately depends on how increased AI capabilities shape human beliefs and collaborative efforts.
Isotropy Cliffs: The Geometric Signature of Decision-Making in Large Language Models
各向同性悬崖：大型语言模型决策的几何特征
Coskun, Okan S., Rottach, Florian, Eickhoff, Carsten, Rudman, William
Abstract
We investigate the geometry of decision-making in Multiple Choice Question Answering (MCQA) through the lens of isotropy. Analyzing five open-weight models across diverse datasets, we identify decision-critical transition layers characterized by a shift in isotropy, coinciding with a major representational change and the emergence of task-relevant clusters. We demonstrate that this synchronized geometric behavior is strongly correlated with downstream accuracy ($r\approx0.84$), displaying its relevance for successful decision-making. Furthermore, we show that this transition is robust to prompt variations, suggesting that it reflects a general mechanism of model behavior.
Similarity Weighted Aggregation with Global Differential Privacy for Federated Brain Lesion Segmentation
具有全局差分隐私的相似性加权聚合用于联邦脑病灶分割
Khan, Muhammad Irfan, Lehtonen, Eero, Obradovic, Joni, Kontio, Elina, Alhoniemi, Esa, Khan, Suleiman A., Jafaritadi, Mojtaba
Abstract
Federated Learning (FL) enables collaborative training of machine learning models across multiple institutions without sharing sensitive data, making it particularly suitable for medical imaging applications. However, heterogeneous data distributions across institutions and potential information leakage through model updates remain important challenges. In this work, we propose DP-SimAgg, a privacy-preserving federated learning framework that integrates similarity-weighted aggregation with a server-side differential privacy mechanism. The proposed method applies L2 clipping to bound collaborator updates, computes similarity-based aggregation weights to mitigate the effects of non-IID data distributions, and injects calibrated Gaussian noise at the central server, providing per-round privacy guarantees under the assumed sensitivity bound. The framework is implemented using Intel's OpenFL platform and evaluated on the FeTS 2022 dataset consisting of 1251 multi-modal MRI scans for brain tumor segmentation. Experimental results demonstrate that DP-SimAgg maintains competitive segmentation performance while providing privacy protection. Under a strict per-round privacy budget (epsilon = 1, cumulative epsilon_total = 20 over 20 rounds), the method achieves Dice scores of 0.6357, 0.5305, and 0.5274 for the enhancing tumor (ET), tumor core (TC), and whole tumor (WT) regions, respectively. With a more relaxed per-round budget (epsilon = 10, cumulative epsilon_total = 200), performance approaches that of the non-private baseline while incorporating a central Gaussian mechanism with per-round (epsilon, delta)-DP accounting under the assumed sensitivity bound. These results highlight the potential of DP-SimAgg for enabling privacy-preserving collaborative learning in medical imaging applications.
Assuming You Knew: Fixing an Epistemic Semantics for Flow Policies Using Agentic AI
假设你知道：利用自主人工智能修正流政策的认知语义学
Naumann, David A.
Abstract
Many high-level security requirements are about the allowed flow of information in programs and are difficult to make precise because they involve selective downgrading. Notions from epistemic logic have emerged as a good approach to policy semantics but a robust general framework remains elusive. A paper appearing in CSF 2018, entitled ``Assuming You Know: Epistemic Semantics of Relational Annotations for Expressive Flow Policies'', attempted to provide a unifying framework---but the formalization was sketchy and a correction was announced during the conference presentation. With aid from an agentic AI coding assistant, a corrected formalization has been machine checked in the Rocq proof assistant. The simplicity and generality of the framework may help compare different policy specification styles and enforce them by leveraging existing techniques.
This study addresses the Variable Gapped Longest Common Subsequence Problem (VGLCSP), a variant of the classical longest common subsequence problem with additional gap constraints and applications in sequence alignment and time-series analysis. While the two-sequence version has been widely studied using dynamic programming, the generalized multi-sequence form is usually solved with beam search-based heuristics, whose hand-crafted designs often lack robustness. To overcome this limitation, we propose a learning-based approach for automatically designing more effective data-driven heuristics. The heuristics are represented by a neural network with predefined architecture, whose weights are optimized by a genetic algorithm within a neuro-evolutionary framework. The learning process alternates between weight optimization and evaluation within an iterative multi-source beam search procedure, a state-of-the-art method for the problem. Rather than constructing solutions directly, the neural network learns to guide the search process, producing a neuro-evolved heuristic. We further introduce an ensemble heuristic that combines the scores of learned and the best-performing hand-crafted heuristic. Integrated into the iterative multi-source beam search framework, the resulting hybrid approach outperforms existing methods on both synthetic benchmark instances and newly introduced real-world instances with data-driven gap constraints.
Large language models have made it possible to generate executable computer-aided design (CAD) programs from natural-language descriptions or images. However, existing methods represent modeling processes as backend-specific sequential scripts with implicit dependencies or as static geometry, making it difficult to simultaneously preserve construction history, stable topological references, and feature-level editability across different CAD systems. We present CADIR, an agent-friendly executable intermediate representation for CAD generation and cross-backend editing. Built on the OCCT geometry kernel via OCP, CADIR provides explicit, compositional modeling operations and fine-grained execution diagnostics. During program execution, CADIR records modeling operations, parameter dependencies, constraints, and topology selections in a construction graph. To enable reliable cross-backend reconstruction, we introduce Geometric Signature Matching, which identifies corresponding edges and faces despite parameter changes and backend differences, allowing adapters to reconstruct native editable feature histories in FreeCAD, SolidWorks, and Fusion 360. Building on this representation, we further propose a construction-graph retrieval method for text and image queries that supports both full-graph and subgraph retrieval, enabling agents to leverage complete models and modeling substructures. Extensive experiments demonstrate that CADIR achieves higher geometric fidelity and execution reliability than existing CAD representations, that construction-graph retrieval further improves model generation quality, and that cross-backend editing enables reliable model reconstruction and post-reconstruction editing across multiple CAD environments.
Modeling Social Dynamics with an LLM-Enabled Agent Based Network-Dynamic (LAND) Model
基于 LLM 的代理网络动态模型（LAND）建模社会动态
Ng, Lynnette Hui Xian, Carley, Kathleen M.
Abstract
Social dynamics encode the process in which individual network and discourse interactions aggregate into collective influence, narrative dominance and coordinate behavior. This paper uses the the GhostField architecture, a hybrid LLM-Enabled Agent Based Network-Dynamic (LAND) model as a social simulation framework to build the AuraSight scenario. In the AuraSight scenario, 314,244 heterogeneous cyber social agents and human actors exchange 529,327 messages over 30 days surrounding a fictional international song-writing contest. We methodologically examine emergent social dynamics across four analytical layers: ego-network topology, semantic network evolution, coordination dynamics and influence dynamics. Our results show how generated social simulations do also produce social dynamics, and how the dynamics of coordination and influence emerge not from individual agents but from the recursive interaction between network topology and narrative exchange.
Long-term memory is essential for LVLM agents to maintain consistency and integrate information across extended multimodal interactions. Existing agent memory systems, however, often reduce visual experiences into textual summaries or rely on static retrieve-then-reason pipelines, which are inefficient at query time and brittle when questions require image-text binding, temporal updates, or visual details. We propose Prospective Multimodal Memory Compilation, a framework that shifts part of the memory reasoning process from query time to memory consolidation time. Given accumulated multimodal interactions, a Questioner predicts future question candidates, a Planner compiles question-conditioned multimodal memory programs, and a Doubter verifies whether the planned evidence path can support the predicted answer. The verified question-program pairs form a structured question bank for efficient query-time routing and evidence retrieval. Experiments on multimodal long-term memory benchmarks show that our method improves answer quality and visual evidence recall while reducing query-time token and latency costs. Extensive ablations analyze the effects of self-feedback, dynamic planning, raw-image access, and question bank coverage.
Large language model agents have shown strong capabilities in generating coherent and contextually appropriate responses, yet robust long-horizon dialogue remains limited by the lack of external memory that is traceable, updatable, and diagnostically transparent. Existing memory-augmented agents often store memories as isolated records or overwritable states, making it difficult to preserve how information originates, evolves, conflicts, or becomes obsolete over time. We propose TrajWiki, a trajectory-based memory framework for long-horizon conversational agents. Instead of treating memory as static entries, TrajWiki represents each memory as a source-grounded evolution trajectory, maintained through immutable episodic snapshots and claim-level operations such as ADD, REVISE, and DEPRECATE. To reduce fragmentation and retrieval cost, TrajWiki further introduces Memory Wiki, a persistent intermediate layer that incrementally compiles dialogue history into structured and interlinked wiki pages capturing salient entities, events, quantities, topics, and conflicts. At inference time, queries are routed hierarchically from relevant wiki pages to linked memory trajectories, then to corresponding snapshots and source messages for evidence-grounded answer synthesis. Experiments on LoCoMo and MedMT show that TrajWiki improves long-horizon dialogue performance across both open-source and closed-source LLM backbones, while providing greater interpretability and diagnostic visibility into memory evolution, retrieval failures, and answer generation.
Existing search-augmented LLM agents are trained using Reinforcement Learning to boost its reasoning capabilities. However, these approaches primarily rely on outcome-level rewards, which provide little supervision over search behavior and overlook agent's ability to decompose complex queries properly. To mitigate this issue, we propose PROGRESS which utilizes teacher-guided coverage reward to explicitly shape decomposed query generation of the policy model. During training, frozen teacher models are used to decompose complex queries into essential search queries. These essential search queries are utilized to guide the search behavior of the policy model. Integrated into an R1-style training framework, our approach provides lightweight guidance over query decomposition decisions without dense process-level supervision. Experiments show that coverage-guided RL improves overall task performance, highlighting the importance of explicitly supervising search strategies in agentic LLMs.
The effective use of search engines by large language models (LLMs) remains a significant challenge, particularly in complex, multi-hop question-answering (MHQA) tasks. These tasks require the model to decompose questions into subqueries, retrieve relevant information, and synthesize answers from multiple sources, often leading to cascading errors due to poor retrieval in early stages. Reinforcement learning (RL) has shown promise in improving LLMs' search capabilities, but it often suffers from sparse rewards during training, hindering the model's ability to learn effectively. To address these challenges, we introduce Guided Retrieval Training (GRT), a novel method that improves the performance of a search agent by restricting the retrieval process during RL training using ground truth information. By focusing on a curated set of relevant documents, GRT provides the model with a stronger learning signal, mitigating the problem of sparse rewards and improving its ability to generate accurate subqueries and synthesize correct answers. Our experimental results demonstrate that GRT achieves consistent performance improvements over existing methods, such as Search-R1, across a wide range of question-answering (QA) tasks. Notably, GRT excels in MHQA tasks, achieving over 40% improvements in performance. Additionally, GRT enhances training efficiency by achieving better QA performance with fewer training steps.
Passing Coarse Marginal Checks Can Be Cheap: Persona Mixtures and Imprecise Treatment-Response Estimates in an LLM Persona Panel
通过粗略边际检查可能是廉价的：LLM角色面板中的角色混合与不精确的治疗反应估计
Nakajima, Yohei
Abstract
Large language models are increasingly used as synthetic research participants and are often validated by whether their marginal responses resemble human data. We study a fixed panel of sixteen lightweight persona-conditioned GPT-4.1 configurations in repeated strategic games. The panel met preregistered broad-reference condition-mean criteria in three of four repeated-game cells; the sole miss was 0.011 below the lower reference bound. Variation was strongly prompt-indexed, but its share depended on uncertainty assumptions: fixed-panel symmetric-Dirichlet sensitivities produced median between-prompt shares of 63%-71% under Jeffreys alpha=0.5 and 47%-53% under alpha=1, while finite-opportunity plug-in estimates were 85%-96%. Aggregate continuation-probability contrasts were +0.083 and +0.078, with conservative simultaneous 95% intervals [-0.171, +0.330] and [-0.181, +0.330]. The treatment jointly changed the continuation process and its textual representation. A separate wording-and-position operation shifted cooperation from 0/40 to 37/40 in the bare configuration, and a label conflict also revealed representation control. The original persona-level p13 result was not prospectively family-controlled, while a post-adjudication exact gate was structurally underpowered; p13 is therefore a replication target rather than a finding. External review exposed family-error, dependence, construct, and boundary-uncertainty defects, and zero-call reanalysis changed the interpretation without rewriting the historical record. The registered marginal criteria could be passed without precisely estimating the treatment-response object. A public capsule verifies 4,916 confirmatory Phase 3-5 runs with no live model calls. The results concern one fixed model-prompt panel and do not establish human substitutability.
Auditing Discovery Claims: A Two-Sided Criterion for Agentic Science, with the Negative Side Decidable
审计发现声明：一种双面标准用于代理科学，其负面方面是可判定的
Chen, Wenhui, Chen, Jianlin, Lin, Ziyao, Vong, Chi Man
Abstract
When a self-improving AI-for-science system claims a new capability, the evidence is usually a benchmark delta, a description-length gate, or a p-value. None separates a real gain from extra search, from a changed verifier, or from adaptation to a fallible oracle. We build a two-sided audit whose negative side is a formal fact: a pseudoknot-free oracle provably cannot represent a crossing base pair, so the prior verifier's range is bounded exactly, offline, before any run. "New" is relative to the agent's prior self, never to the base model. First, how far a single fallible oracle can inflate a capability claim. An invented, solver-free operator solves 43/60 crossing RNA targets under the predictor it optimizes, above a context-free floor of 0/60; under three predictors, 1/60 survives. Paired on the same 43 targets, a predictor the operator never saw confirms 2 of its designs against 26 for a minimum-free-energy solver (p = 8e-7). No statistic computed from the system and its own oracle sees that gap. Second, agent-written procedures can beat a human-written one under a judge no objective can flatter, at a fraction of the compute. Of six frontier models, the two whose operators ran without timeouts carry over at 0.293 against our 0.095 (n = 951 paired units, target-clustered [+0.108, +0.297], p = 5e-5) while spending 4.6-10x fewer oracle calls. Three rungs: difference under an outside adjudicator (reached), not bought with compute (reached, both directions), mechanism identified and transferable (not reached; seven candidates tested, none moves the statistic). The ceiling is the panel itself: its three predictors share nearest-neighbour thermodynamic parameters, two agreeing at kappa = 0.673. The audit is as unsparing about our own system: matched undirected search is an exact zero, and a search-free probe puts 84% of our headline effect on targets a random sequence already solves.
SCHEDBench: A Benchmark for Evaluating LLM Constraint Faithfulness in Natural-Language Combinatorial Scheduling
SCHEDBench：评估大型语言模型在自然语言组合调度中约束忠实性的基准测试
Sharma, Shrenil Shaun, Sharma, Avi
Abstract
This paper introduces SCHEDBench, a natural-language benchmark for evaluating combinatorial scheduling constraint faithfulness under surface-form variation. Grounded in canonical scheduling instances and solver-derived feasibility and optimality, SCHEDBench assesses whether large language models (LLMs) generate schedules with the same constraint-feasible behavior across varied natural-language (NL) surface forms. SCHEDBench spans 1,132 instances across job-shop scheduling problems (JSP), single and multi-mode resource-constrained project scheduling problems (RCPSP), nurse rostering/scheduling, and curriculum timetabling problems of varying difficulty. Instances are templated into natural language problems using domain-specific templates, themed entities, lexical-syntactic template rephrasing, and constraint-level surface-form variation, with reference solutions verified for feasibility and objective optimality. Across thirteen frontier and open-weight LLMs, we find that models are not reliably invariant to semantically equivalent renderings of the same scheduling problem. Surface-form variation reduces feasibility and induces above-noise shifts in per-instance hard-constraint violations on matched instances. Among the tested isolated axes, constraint reordering yields the clearest above-noise sensitivity.
Judging Is Not Enumerating: Silent Omissions in LLM-Authored Acceptable Sets
判断并非枚举：大型语言模型创作的可接受集合中的静默遗漏
Chen, Wenhui, Chen, Jianlin, Lin, Ziyao, Long, Peiji, Vong, Chi Man
Abstract
Language models are increasingly promoted from examinees to examiners: they write the test suites, answer keys, rubrics, and reward functions that define correctness for other systems. We measure the capability that role assumes and find it lacking under the protocol the role is usually deployed with, one-shot greedy authoring with no test-time reasoning. Across four reference constructions - two with complete finite truth, one with a hardened executable reference (HumanEval+/MBPP+), one with an explicitly incomplete lexical reference (WordNet) - models judge whether a candidate belongs far better than they author the set itself. On the incompleteness-proof algorithmic construction the gap is +0.34 to +0.29 F1 over a 24x parameter range and does not close; on executable code, models judging at F1 0.74-0.90 author suites admitting only 19-42% of oracle-correct solutions. A control locates the deficit: asked to emit the predicate rather than its extension, the same models reach F1 about 0.99. The failure is not missing knowledge or an inability to specify, but an inability to materialise the region a specification induces. The dominant error is omission, which resists audit: an over-inclusion is a token a reviewer can challenge, a missing member an absence whose discovery is the authoring problem itself. Models detect planted over-inclusions 6-7x more often than planted omissions, and a production deployment of 43,227 items fails omission-first at 10:1. Wired into RLVR, an authored key costs 1.9 points of accuracy against an exact oracle and 18.5 WordNet-relative (six paired seeds, p=0.031). Gating authored verifiers on a known-correct probe cuts false rejection from 58-92% to at most 5%, but keeps only 5-39% of suites. Repairing them instead, by rewriting each wrong expected value to what a reference execution returns, raises yield 3.3-10.6x across four author families.
The emergence of Agentic AI systems, characterized by autonomous reasoning, multi-agent collaboration, tool orchestration, adaptive decision-making, and persistent memory, represents a fundamental shift from traditional AI pipelines to dynamic software ecosystems. While AI Technical Debt (AITD) has been widely studied in machine learning and software engineering, existing models assume static, component-level architectures and fail to capture the dynamic and emergent behaviors of agentic environments. To address this gap, this paper introduces Agentic Technical Debt (AgTD), defined as technical debt that emerges, accumulates, propagates, and amplifies due to the autonomous and collaborative nature of Agentic AI systems. Building on our prior systematic scoping review of 31 AITDs across seven root-cause categories, we employ a theory-informed transformation methodology to reinterpret these debts in Agentic AI through direct transformation, contextual transformation, and manifestation expansion. We present the first systematic mapping of established AITDs to their agentic manifestations, showing how conventional debts evolve into system-level liabilities, including memory inconsistencies, orchestration fragility, cascading failures, and unsafe autonomous decision-making. Our findings show that technical debt extends beyond software artifacts to encompass agent behaviors, coordination mechanisms, and interactions among agents, tools, and execution environments. We further examine its implications for AI Trust, Risk, and Security Management (AI TRiSM), highlighting impacts on trustworthiness, governance, security, operational resilience, and Sustainability Technical Debt. Overall, this work establishes AgTD as a foundational software engineering construct and provides a transformation framework, taxonomy, and research agenda for managing technical debt in autonomous multi-agent AI systems.
Multimodal large language models (MLLMs) exhibit strong vision--language capabilities but may also memorize and disclose sensitive information. Machine unlearning seeks to remove designated knowledge without retraining from scratch while preserving general utility. Existing privacy-oriented benchmarks primarily adopt profile-level deletion, whereas practical requests are often finer grained: a model should forget a specified attribute while retaining non-sensitive information about the same identity. We therefore introduce attribute-level MLLM unlearning as a finer-grained task and construct a benchmark spanning long-text, numeric, and short-text targets, multiple forget ratios, and diverse question types. Our evaluation reveals that target and retained attributes share identity-specific and visual evidence, making selective forgetting susceptible to residual leakage or collateral degradation; accordingly, existing methods exhibit unstable forgetting--retention trade-offs in this setting. To address this challenge, we propose Causal Localization and Retain-Aware Projection (CLRP), a lightweight training-free framework. CLRP uses activation patching to identify the layer that causally mediates target-attribute disclosure, then applies a retain-aware projection that removes the target-attribute subspace while preserving same-identity evidence. Experiments across multiple widely used MLLMs with distinct architectures and parameter scales demonstrate the effectiveness of CLRP.
Chinese Translation
多模态大型语言模型（MLLMs）展现出强大的视觉-语言能力，但也可能记忆并泄露敏感信息。机器遗忘旨在在不从头重新训练的情况下移除指定知识，同时保持一般效用。现有的隐私导向基准主要采用个人资料级别的删除，而实际请求往往更为细粒度：模型应忘记特定属性，同时保留关于同一身份的非敏感信息。因此，我们引入属性级别的MLLM遗忘作为一项更细粒度的任务，并构建了一个基准，涵盖长文本、数字和短文本目标、多种遗忘比例以及多样化的问题类型。我们的评估揭示了目标属性和保留属性共享身份特定和视觉证据，使得选择性遗忘容易受到残余泄漏或附带退化的影响；因此，现有方法在这种情况下表现出不稳定的遗忘-保留权衡。为了解决这一挑战，我们提出了因果定位与保留感知投影（Causal Localization and Retain-Aware Projection, CLRP），这是一个轻量级的无训练框架。CLRP使用激活修补来识别因果介导目标属性泄露的层，然后应用保留感知投影，移除目标属性子空间，同时保留同一身份的证据。针对多种具有不同架构和参数规模的广泛使用的MLLMs的实验表明，CLRP的有效性。
World models have attracted significant attention for their ability to capture and predict the structure and dynamics of the physical world. In this emerging landscape, Joint Embedding Predictive Architectures (JEPA) offer a particularly compelling direction. We study a largely unexplored regime: populous, crowded, and chaotic Global South urban environments, which we call DENSEWORLD. Unlike the lower-density, lane-structured settings that dominate existing evaluations, these scenes exhibit soft spatial boundaries, extreme agent heterogeneity, persistent occlusion, and rapid social negotiation under mixed traffic. We introduce the first large-scale dataset for this regime: 1,000 hours of drive-through, walk-through, and aerial video across 22 cities. Existing JEPA formulations struggle to preserve dense interaction dynamics under heterogeneity and partial observability. We introduce FactorJEPA, which makes world structure a first-class predictive primitive. Rather than encoding the future in a monolithic latent, it composes layout, entities, and interactions, using a visibility gate and separated subspaces to preserve partially observed agents and discourage cross-factor shortcuts. FactorJEPA improves (i) future-latent accuracy (Future-frame L1), (ii) intervention-sensitive prediction (Causal L1), and (iii) robustness to reduced visual evidence (Mask-ratio slope), while exposing (iv) a reproducible motion-information trade-off (Motion cosine). Method rankings replicate across 2B and 1B V-JEPA 2.1 backbones, with rho = 0.895 to 0.978. We publicly release the DENSEWORLD-115k dataset (https://huggingface.co/datasets/anonymousML123/denseworld-115k) and the surgery-trained FactorJEPA checkpoints (https://huggingface.co/datasets/anonymousML123/factorjepa-outputs/tree/main/outputs/full/vjepa_2_1_vitg_1B/train/m09c_surgery_3stage_DI_diheavy_encoder).
Production LLM agents that select from large skill libraries face a limitation that semantic relevance alone cannot resolve: a skill may match a user's topic yet be impossible to execute in the current account state. We present a deployed three-stage selection pipeline for Helpmate, Wix's customer-care assistant. First, a recall-oriented semantic matcher identifies messages related to a ten-skill domain family without consulting account state. Second, a deterministic executability gate removes candidates whose internal hard-stop conditions hold. Because the gate and the skill evaluate the same exit predicates, every blocked candidate would be unable to complete under the same account state, provided predicate parity is preserved and both checks observe fresh authoritative state. Finally, the LLM decides whether to invoke one of the remaining candidates. In a post-launch production analysis of 756.6K user messages across 267.6K conversations, semantic matching retained 174,927 messages (23.1%). Within this matched stream, the gate removed 1,039,462 of 1,749,270 skill-message pairs (59.4%), saving 228.8 million skill-description tokens -- 59.1% of the post-semantic skill-description footprint. Together, semantic matching and executability gating reduced skill-description context by 90.5% relative to exposing all ten skills to every message. To test whether this pruning affects model behavior rather than context size alone, we replayed a risk-enriched cohort of 1,000 conversations with all ten skills exposed. The model selected a production-blocked skill in 78 conversations (7.8%). This counterfactual result shows that deterministic gating prevents non-executable candidates from influencing model selection, while not claiming downstream tool execution or customer-outcome effects.
Control Under Compression: Reliability Frontiers for Tool-Using Agents
压缩下的控制：工具使用代理的可靠性前沿
Hou, Yinghan, Yang, Zongyou
Abstract
Tool-using language-model agents are governed not only by task prompts but also by persistent system-side instructions that specify tools, arguments, policies, execution protocols, and recovery. Compressing these agent control contexts (ACCs) can reduce input cost and context use, yet existing prompt-compression evaluations do not reveal whether the resulting control remains operationally reliable. We introduce CompressAgent, an environment-verified benchmark for ACC compression across nine independently constructed ACCs, three task families, three fixed Qwen API model identifiers, six retained-context budgets, and 15,525 runs. We uncover a nonlinear, method-dependent reliability frontier. At 75% retained context, generic rewriting and section-based compression achieve 92.7% and 92.4% success, close to the 93.8% full-context baseline. Between 50% and 35%, methods diverge sharply; at 35%, section-based, obligation-aware, and generic rewriting achieve 47.0%, 39.0%, and 19.9%. At retained-context budgets from 25% to 10%, executable protocols become fragile. Reliability also varies substantially across ACCs, making universal compressor rankings inappropriate and motivating per-context qualification. Failure analysis shows that compression primarily surfaces as tool-execution and action-parsing errors. These findings recast ACC compression from token reduction into a runtime-reliability problem that must be evaluated through executable outcomes.
Role-Decoupled Attention Residuals: Separating Matching and Content Retrieval Across Depth
角色解耦注意残差：在深度之间分离匹配与内容检索
Wang, Kehan
Abstract
Depth-routing residual architectures allow Transformer layers to retrieve earlier representations instead of inheriting only the immediately preceding state. Existing Block Attention Residuals, however, use a single content-dependent depth mixture to construct the inputs to queries, keys, and values. This design couples two functionally different decisions: queries and keys determine where attention matches, whereas values determine what content is retrieved. We therefore ask whether matching and content retrieval should be forced to read from the same depth. We introduce Role-Decoupled Attention Residuals (RD-AttnRes), a minimal extension that shares one depth route between queries and keys while learning an independent value route over the same residual sources. Tying the two routing queries exactly recovers the parent architecture, while decoupling them adds only one model-width vector per layer and introduces no additional token-to-token attention operation. We evaluate RD-AttnRes using a frozen, paired pretraining protocol on FineWeb-Edu with five matched seeds for both 120M- and 343M-parameter models and a 2.0B-token training budget. RD-AttnRes improves validation negative log-likelihood in all 10 matched comparisons. The mean reductions are 0.0301 and 0.0247, corresponding to perplexity reductions of 2.97 percent and 2.43 percent at 120M and 343M parameters, respectively. Early-budget controls indicate that neither the additional parameter count, duplicated routing execution, nor a fixed value route reproduces the improvement. Routing diagnostics further reveal persistent divergence between the query-key and value depth distributions. These results suggest that, within the evaluated training regime, attention matching and content retrieval benefit from distinct reads over the residual hierarchy.
Inter-Residue Geometry Attention for Antibody-Specific Epitope Prediction
针对抗体特异性表位预测的残基间几何注意力
Fan, Chuanliu, Yu, Nan, Wu, Junjie, Fu, Guohong
Abstract
Antibody-specific epitope prediction aims to identify which antigen residues are recognized by a given antibody, a task that depends on the three-dimensional complementarity between antibody CDRs and the antigen surface. Existing methods usually leverage PLM embeddings and inject structure through additional graph, surface, or point-cloud encoders, where the positional mechanism inside attention remains largely tied to one-dimensional sequence order. For proteins, the analogue of a token offset is not only sequence separation, but also the three-dimensional displacement between residues after folding. This raises a question, can folded residue geometry serve as the positional mechanism of attention itself? We propose Local-Frame 3D Rotary Position Encoding (LF3DRoPE), which expresses inter-residue displacements in backbone-defined local frames and injects them directly into rotary attention. This design preserves continuous directional geometry while ensuring invariance to global $\mathrm{SE}(3)$ transformations. On the AsEP benchmark, LF3DRoPE achieves state-of-the-art $\mathrm{MCC}$ on both ratio and epitope-group splits. Ablations and rigid transformation tests show that local three-dimensional geometry provides information beyond sequence-order attention while preserving invariance to arbitrary global coordinate systems. Mutation ranking results further indicate that LF3DRoPE captures antigen-specific structural compatibility.
Chinese Translation
抗体特异性表位预测旨在识别特定抗体所识别的抗原残基，这一任务依赖于抗体互补决定区（CDR）与抗原表面之间的三维互补性。现有方法通常利用预训练语言模型（PLM）嵌入，并通过额外的图形、表面或点云编码器注入结构，其中注意力机制内部的位置信息在很大程度上仍然与一维序列顺序相关。对于蛋白质而言，标记偏移的类比不仅是序列分离，还有折叠后残基之间的三维位移。这引发了一个问题：折叠后的残基几何是否可以作为注意力机制本身的位置信息？我们提出了局部框架三维旋转位置编码（Local-Frame 3D Rotary Position Encoding, LF3DRoPE），该方法在主链定义的局部框架中表达残基间的位移，并将其直接注入旋转注意力中。该设计在保持连续方向几何的同时，确保对全球 $ ext{SE}(3)$ 变换的不变性。在AsEP基准测试中，LF3DRoPE在比例和表位组拆分上均实现了最先进的 $ ext{MCC}$。消融实验和刚性变换测试表明，局部三维几何提供了超越序列顺序注意力的信息，同时保持对任意全球坐标系统的不变性。突变排名结果进一步表明，LF3DRoPE捕捉了抗原特异性的结构兼容性。
Fighting Fire with Fire: On the Feasibility of Protecting Exercises Against AI Cheating
以火攻火：保护教育练习免受人工智能作弊的可行性研究
Braun, Tobias, Grebe, Jonas, Rethfeld, Louis, Rohrbach, Marcus
Abstract
The widespread adoption of generative AI enables students to outsource cognitive effort to increasingly capable assistants, creating an illusion of competence while undermining the independent reasoning that education aims to cultivate. We investigate whether adversarial machine learning can be repurposed to protect educational exercises against such corrosive reliance. Our approach uses multimodal multiple-choice questions whose visual components can be protected with subtle visual perturbations that steer AI solvers toward designated incorrect answers. These responses form a statistical fingerprint: students who blindly copy a solver reproduce the induced answer pattern more frequently than genuine students. We study the feasibility of this paradigm under realistic black-box assistant assumptions using three of the most common state-of-the-art multimodal language models: Anthropic's Claude, Google's Gemini, and OpenAI's ChatGPT. By using accessible surrogate models, we optimize adversarial perturbations that induce consistent response patterns. Those patterns enable principled detection through statistical hypothesis testing. These findings establish both the promise and the limitations of fighting machine-assisted reasoning with the vulnerabilities of the machines themselves.
In post-disaster scenarios, unmanned aerial vehicles (UAVs) are critical for establishing emergency communication networks. For time-critical rescue missions, information freshness is crucial because decisions based on outdated data may lead to ineffective control actions. This paper investigates age of information (AoI) minimization for UAV-assisted emergency communications with heterogeneous emergency services. We model bursty packet arrivals using a Markov-modulated Poisson process and adopt finite blocklength theory to capture the coupling among transmission duration, packet completion, and AoI evolution. To balance delay-tolerant long-packet transmission and urgent short-packet response, we propose a mini-slot-embedded scheduling mechanism with adaptive checkpoint-interval selection. We formulate the joint optimization of UAV trajectory control, user scheduling, and checkpoint-interval selection as a multi-agent decision problem, and develop MA-HEAD-Net, an adaptive rule-guided multi-agent deep reinforcement learning framework. MA-HEAD-Net incorporates communication-domain rule priors into a gated multi-head policy, where adaptive gates regulate the contributions of rule-prior and learned-policy logits for different subtasks. The policy and gating components are jointly optimized under multi-agent proximal policy optimization. Simulation results show that MA-HEAD-Net improves policy-formation efficiency compared with representative multi-agent deep reinforcement learning baselines and achieves lower AoI than both learning-based and heuristic methods in dynamic UAV-assisted emergency communication scenarios.
Lifelong LLM agents increasingly adapt through external learning states that store past interactions as retrievable memories or reusable skills, yet existing benchmarks rarely account for how the path of accumulated experience shapes what agents transfer and retain. In this work, we establish PATH-Bench, a benchmark for path-dependent evaluation of lifelong agents. PATH-Bench estimates directed task relationships via multi-model in-context learning, constructs probe-centered sequences with controlled helpful and interfering histories, and repeatedly evaluates probe tasks to measure average performance, forward transfer, backward transfer, and forgetting. We evaluate eight representative agents on single-turn code generation and multi-turn tool-use tasks under positive- and negative-dominant histories. Benchmark results show that experience utility depends jointly on how experience is represented and on the task's interaction structure, that strong transfer does not ensure retention, and that later experience can reshape gains acquired earlier in the learning path. Based on these findings, we propose Selective Experience Use (SEU), an agent harness that regulates how path-accumulated experience influences each new task, admitting helpful items while filtering out potential interference. SEU consistently reduces forgetting while improving forward transfer in the majority of settings. The PATH-Bench provides both a controlled evaluation framework and actionable guidance for designing more selective and robust lifelong agents.
The Graph Language: How Knowledge Graphs Speak to Large Language Models
图谱语言：知识图谱如何与大型语言模型对话
Pirrò, Giuseppe
Abstract
Large Language Models (LLMs) excel at reasoning but benefit from grounding provided by Knowledge Graphs (KGs). However, integrating these paradigms is challenging. We introduce GRALAN, which enables KGs to speak directly in the LLM's semantic space through relational tokens that preserve graph structure. GRALAN-s trainable language mediator generates structured tokens for any frozen LLM, creating a foundation for knowledge-intensive applications. We demonstrate its effectiveness in question-answering by re-framing the task as entity classification over question-focused subgraphs. Experiments show that GRALAN significantly outperforms existing methods, particularly on complex multi-hop reasoning tasks, establishing a new paradigm for KG-LLM integration that maintains structural fidelity while leveraging LLMs' reasoning capabilities.
Co-evolution of social reward and punishment under institutional interventions
制度干预下社会奖励与惩罚的共同演化
Nguyen, Van An, Huynh, Vuong Khang, Nguyen, Hoai Thuong, Duong, Duc Tin, Gia, An Nguyen, Nguyen, Tat Kien, Bui, Huu Loi, Tra, My Nguyen, Duong, Ho Nam, Phan, Ba Thanh, Vo, Thanh, Hoang, Dinh Anh Trung, Bashir, Adeela, Song, Zhao, Duong, Manh Hong, Trang, Le Hong, Han, The Anh
Abstract
We investigate how peer and institutional incentives jointly shape the evolution of cooperation, social welfare, and enforcement efficiency in social dilemmas. In a Prisoners Dilemma with four strategies, unconditional cooperators (C), defectors (D), social punishers (SP), and social rewarders (SR), we allow decentralised peer incentives and centralised institutional incentives to act simultaneously, with the institution able to reward or punish any subset of strategies. In infinite well-mixed populations, we analyse the resulting four-strategy replicator dynamics, and in structured populations we use agent-based simulations on square lattices to study spatial effects and network reciprocity. Intervention schemes are evaluated by equilibrium states and evolutionary flow for infinite well-mixed populations, by cooperation levels and social welfare for structured populations, defined as aggregate population payoff net of institutional cost. We find that peer punishment most strongly promotes cooperation, whereas peer reward is more beneficial for social welfare. Institutionally rewarding peer incentive strategies substantially improves both cooperation and welfare, while subsidising unconditional cooperators has little impact. Under institutional punishment, directly penalising defectors is the only consistently effective policy; punishing peer incentive strategies dismantles decentralised incentives, reduces cooperation, and harms social welfare, showing that maximising cooperation does not necessarily optimise overall societal benefit. Our findings provide design principles for institutions seeking to balance cooperation promotion with welfare maximisation.
Humans Are More Diverse: Frontier LLMs Show Extreme Policies in Idealised AI Development Races
人类更具多样性：前沿大型语言模型在理想化人工智能发展竞赛中表现出极端政策
Pham, Phu Hoa, Sy, Duy Minh Dao, Huynh, Trung Kiet, Lam, Phu Quy Nguyen, Tran, Chi Nguyen, Le, Minh Trung, Le, Phong Hao, Nguyen, Dinh Nam, Dong, Thien Ky Nguyen, Domingos, Elias Fernandez, Trang, Le Hong, Han, The Anh
Abstract
An AI development race creates a multi-agent safety dilemma. Each company can develop slowly and safely, or move faster while taking a risk that may remove its final reward. We use this repeated game to study strategic safety behaviour among large language model (LLM) agents in races with two to five players. However, a valid action does not show that an agent understands the game. We therefore place an audit gate before behavioural interpretation. We first verify the game engine, then test rule recall, state tracking, payoff calculation, and stability under different but equivalent task descriptions. We then compare LLM action sequences with an evolutionary game-theory benchmark and published human data, and explore differences across models, risk conditions, personas, and two- to five-player races. The audit shows that strong rule recall can coexist with weak state tracking and expected-payoff calculation. Providing verified arithmetic and changing the response representation can also change later actions, even when the game rules stay fixed. Across seven tested model endpoints, aggregate rates hide large differences in action sequences, responses to opponents, and responses to race position. Patterns across the tested three- to five-player races are also model-specific rather than a single effect of adding competitors. These results show why multi-agent AI-race simulations need validity checks and trajectory-level analysis before their outputs are described as strategic, human-like, or safety-aware. Our findings are exploratory and apply only to the tested models, prompts, and decoding settings.
Reputation-driven Cooperation in Lattice-based Decentralized Federated Learning through Evolutionary Game Theory
基于声誉驱动的格子结构去中心化联邦学习中的合作研究：演化博弈理论视角
Huynh, Phuc Hoang Truong, Vinh, Dung Tran, Lam, Khoa Duc Anh, Truong, An Nghiem Nguyen, Bui, Uyen Nha Tran, Dinh, Khang Nguyen, Gia, Bao Nguyen Le, Nhat, Minh Le Nguyen, Duong, Manh Hong, Han, The Anh, Nguyen, Thi Ai Thao, Trang, and Le Hong
Abstract
Decentralized Federated Learning (DFL) has emerged as an optimal privacy-preserving solution; however, it remains vulnerable to opportunistic behaviors due to the absence of a central coordinator. While Evolutionary Game Theory (EGT) serves as a powerful framework for analyzing such behaviors, existing studies often assume that agents possess perfect rationality and maintain static strategies. To address these limitations, this paper proposes a novel EGT framework designed to analyze strategic evolution and enhance overall system performance. The primary contributions of this work are threefold: First, we model peer-to-peer (P2P) interactions on a lattice network structure under the assumption of bounded rationality. Second, we formulate a comprehensive payoff matrix incorporating training costs, communication overhead, and cooperative rewards, while tailoring a strategy update rule that captures spatial propagation dynamics. Third, we integrate a reputation-based reward-and-punishment mechanism to effectively deter free-riding behaviors. Simulation results demonstrate that the framework significantly outperforms the baseline. Specifically, it increases average accuracy from approximately 70% to 82%, elevates cooperation frequency to approach 100% (compared to below 5% in the baseline), and drops accuracy variance from around 0.40 to 0.002, thereby accelerating uniform convergence and ensuring system stability.
Perspectives on Tsallis Statistics for Artificial Intelligence
关于人工智能的Tsallis统计学视角
da Costa, Kleyton, Modenesi, Bernardo
Abstract
Tsallis statistics generalizes Boltzmann-Gibbs statistical mechanics through a single real parameter $q$ that controls the weight assigned to rare and frequent events. Originally proposed to describe physical systems with long-range correlations, multifractal geometry, and heavy-tailed fluctuations, the framework has become a recurring ingredient in modern artificial intelligence (AI): it underlies sparse attention mechanisms (\textsc{sparsemax} and $\alpha$-\textsc{entmax}), maximum-entropy reinforcement learning with controllable exploration, robust and heavy-tailed probabilistic models, and a family of generalized loss functions and regularizers. This paper offers a structured perspective on where Tsallis statistics meets AI. We first review the mathematical core: $q$-entropy and its variational (maximum-entropy) foundation, the $q$-exponential and $q$-logarithm, the $q$-central limit theorem, $q$-Gaussian distributions, and their dynamical origin in superstatistics, emphasizing the properties that matter for machine learning. We then survey applications across softmax generalization, reinforcement learning, sequential and graph neural models, generative and probabilistic modeling, loss design, and optimization, extracting the recurring design pattern in each case: a tunable interpolation between dense/uniform and sparse/peaked behavior governed by $q$. We further argue that the heavy-tailed weight spectra and gradient-noise statistics empirically observed in deep networks are themselves nonextensive signatures, placing modern learning dynamics within the scope of $q$-statistics. Finally, we discuss methodological pitfalls, the relationship to information geometry and $q$-exponential families, and open directions, arguing that $q$ should be treated as a learnable inductive bias rather than a fixed hyperparameter.
Heterogeneous computed tomography (CT) acquisitions and diverse downstream task requirements limit the transferability of fixed data preparation workflows across data sources and tasks. Existing approaches typically rely on manually designed or dataset-specific rules, making it difficult to accommodate changes in acquisition conditions and analytical objectives without manual intervention. Large language model (LLM)-based agents have shown promise for automating medical workflows, yet their potential for adaptive CT data preparation remains largely unexplored. To bridge this gap, we propose CT-PrepAgent, which enables adaptive CT data preparation through a bounded policy and controlled deterministic execution. Deterministic inspection constructs structured data--task profiles, from which a policy decides an eligible DICOM series or predefined preprocessing profile, while the controlled execution flow guards, resolves, executes, and verifies the decision with bounded recovery when enabled and safe quarantine otherwise. Across three public CT segmentation tasks, CT-PrepAgent derived data-task adaptive preprocessing decisions and achieved the highest macro-average Dice. On two private raw-DICOM cohorts, CT-PrepAgent increased verified output yield from 61.7\% to 70.0\% and yielded similar registration metrics on common verified outputs. Controlled fault and replay tests validate bounded recovery, safe quarantine, and policy-free replay under tested fault and drift settings.
Learning What to Remember and What to Internalize in LLM Self-Evolution via Adaptive Memory-Parameter Coordination
通过自适应记忆参数协调学习大型语言模型自我进化中应记住和内化的内容
Ji, Tianyun, Huang, Zhenya, Liu, Jiayu, Liu, Zirui, Su, Yu, Pei, Hongbin
Abstract
Large language model agents increasingly operate in dynamic environments where tool interfaces, APIs, and user requirements change after deployment. Existing self-evolution methods mainly follow two paradigms: harness-based approaches, which externalize feedback into editable memories or skills for rapid adaptation, and parameter-based approaches, which internalize experience into model parameters for deeper capability improvement. However, using either mechanism alone creates a trade-off between flexibility and performance. This paper asks how an agent can coordinate both channels to achieve robust self-evolution. We present COVE, a unified agent self-evolution framework that combines harness-based and parameter-based learning through task-aware routing, stage-aware scheduling, and knowledge optimization. Through this design, COVE treats self-evolution not as indiscriminate accumulation of experience, but as a coordinated process that matches tasks and knowledge types to appropriate learning mechanisms. Experiments across multiple task categories show that COVE outperforms single-channel evolution strategies, demonstrating more robust and efficient improvement under changing environments.
Cognitive Demand Steering for Adaptive Meta-Reasoning in Large Language Models
大型语言模型中的认知需求引导与自适应元推理
Scoville, John, Chen, Shengzhuang, Bang, Yejin, Winzeck, Stefan, Schwarz, Jonathan Richard
Abstract
Recent meta-reasoning frameworks improve LLM reasoning by wrapping chain-of-thought generation in an iterative control loop, allowing more effective backtracking, termination of reasoning loops, and injection of promising reasoning patterns, among other strategy adjustments. Despite promising results, methods often rely on backward-looking reward functions, utilize coarse search actions, or require additional reasoning controller training requiring many-shot supervision. We introduce Cognitive Demand Steering (CDS), a training-free meta-reasoning framework equipped with residual demand assessment: at each step, an LLM-based progress evaluator characterizes the residual reasoning required to arrive at a solution rather than merely evaluating the previous step. This allows a meta-controller to select reasoning interventions comprising both general-purpose exemplars and actions (e.g., general guidance for quantitative reasoning) that directly tackle this forward-looking demand signal. This shift eliminates the need for any trained component while enabling zero-shot transfer across models and tasks with no adaptation. Rather than relying on coarse characterizations, we employ cognitive scales to both design interventions as well as profile initial problem complexity and residual demand signal over 16 dimensions motivated by cognitive science (e.g., attention and scan, learning and abstraction, spatio-physical reasoning), giving the controller a fine-grained vocabulary for diagnosing. Averaged across three frontier LLMs and six reasoning benchmarks, CDS improves accuracy by $21.9\%$ over direct calls and $9\%$ over standard CoT reasoning, with the largest gains on difficult mathematics and coding tasks.
Deep search has become a fundamental capability of large language models (LLMs) for solving open-domain complex tasks. However, existing approaches typically rely on linear sequential reasoning for both trajectory generation and inference, making it difficult to consistently preserve intermediate states and constraints throughout long-horizon multi-hop search. Consequently, they often suffer from context forgetting, search drift, and inefficient exploration. To address these limitations, we propose $\textbf{G-ReAct}$, a reasoning framework for deep search that organizes reasoning as $\textbf{state evolution over a fixed-topology query graph}$. The evolving graph state explicitly tracks search progress and guides subsequent decisions, transforming exploratory search driven by textual history into graph-guided reasoning under explicit constraints. G-ReAct supports both training and inference: it generates high-quality deep-search trajectories for supervised fine-tuning and provides structured guidance for inference-time search without additional fine-tuning. Experiments demonstrate that with only 1.9K generated trajectories for fine-tuning, Qwen3-30B-A3B-Thinking-2507 achieves $52.6\%$ accuracy on BrowseComp-ZH and $79.0\%$ on XBench, outperforming comparable open-source methods trained on substantially larger datasets, including RL-enhanced methods. Furthermore, when applied at inference time, G-ReAct consistently improves the performance of existing strong LLMs on deep-search tasks. We will publicly release all code and model weights.
Programmable-payment protocols such as x402 enable per-request micropayments, but they do not determine which payable service an autonomous agent should buy under a finite wallet. We formulate this buyer-side problem as agent-native payment decision-making: contextual provider selection under wallet pressure, chosen-only paid feedback, and changing market conditions. We propose 402Pilot, a protocol-agnostic buyer-side decision layer between autonomous agents and payment execution that implements purchasing policies for selecting among payable providers. We instantiate it with PA-DCT, a payment-aware discounted contextual Thompson-sampling policy that adapts purchasing decisions under wallet pressure while learning from post-payment feedback. To evaluate buyer-side payment policies, we introduce 402Pilot-Bench, a frozen-replay benchmark spanning 823 tasks, five heterogeneous provider pipelines, and three market regimes, each evaluated over 30 paired seeds. PA-DCT achieves the strongest fixed-wallet adaptive trade-off among non-oracle policies: it maintains competitive service quality while spending only 39 to 43 percent of the wallet and reallocates spending as market conditions change. It attains the best non-oracle PA-gap/T under the price shock and the best mean and worst-case ranks across the nine scenario-metric combinations of quality, ROI, and PA-gap/T. Comparisons with learning baselines and component ablations further support the effectiveness and design of the proposed decision policy. These results suggest that programmable payment must be complemented by buyer-side decision-making capable of learning service value and adapting purchasing decisions accordingly.
Stage-one stellarator design searches a high-dimensional family of three-dimensional plasma boundaries and fixed-boundary MHD equilibria for configurations that jointly meet requirements on confinement, field-line topology, force balance, stability proxies, and geometry. These specifications do not provide a general constructive map to a validated finite-beta equilibrium. High-quality targets are commonly developed through iterative numerical optimization whose outcome depends on the initial configuration, active Fourier resolution, objective priorities, and local solver budget. Coordinating this process is computationally costly and expert-intensive, limiting both design throughput and the production of consistently evaluated data. We present a proof of concept for \emph{agentic} stage-one optimization. A bounded language-model agent diagnoses the current equilibrium and selects the next local optimization experiment, while deterministic DESC execution owns prescribed profiles and flux, symmetry, metric evaluation, solver validity, and acceptance. On a common-budget subset from an expanding finite-beta campaign, the number of gate-valid configurations increases from five inputs to nineteen outputs; median Boozer QS RMS decreases from $2.39\times10^{-4}$ to $1.07\times10^{-4}$, and median maximum principal curvature decreases from $62.56$ to $33.00\,\mathrm{m}^{-1}$. A complementary long route achieves a $9.10\times$ QS reduction while repairing magnetic-well and curvature defects. The system also records every attempted local action as transition evidence, yielding 734 structured parent--action--outcome records in the reported experiments. These results show that agentic outer-loop control can sustain finite-beta, multi-objective search and turn repeated optimization into a scalable source of improved equilibria and reusable decision data.
High-Stakes Decisions with Language Models: Insights from Emergency Triage
语言模型在高风险决策中的应用：来自紧急分诊的启示
Yamin, Khurram, Kelly, Christopher, Wilder, Bryan, Horvitz, Eric
Abstract
High-stakes decisions under uncertainty, such as medical emergency triage, require more than accurate predictions. They depend on estimating the likelihood of alternative outcomes while explicitly weighing the consequences of different actions, principles that have long formed the foundation of medical diagnosis and decision making. Yet language models are increasingly used for high-stakes clinical recommendations without explicit specification of the utilities governing these decisions. Here we show that emergency triage with language models can be understood within a probabilistic decision framework, providing a case study of a broader decision-analytic paradigm for steering, evaluating, and deploying language models in high-stakes settings. Using clinical vignettes from a structured evaluation of a consumer triage system, we analyze recommendations for treatment under alternative utility functions that specify the relative costs of missed emergencies and unnecessary escalation. We find that capable language models adjust recommendations in response to stated utilities, revealing that the same underlying predictions can support markedly different decision policies. These findings show that effective deployment depends not only on improving predictions but also on making decision objectives explicit. More broadly, they suggest that language models for high-stakes applications should be understood and evaluated as probabilistic decision systems whose recommendations depend jointly on predictive performance and explicit utilities.
Constructing an executable chemical-process model remains manually intensive. Chemical engineers translate underspecified requests into coupled decisions about unit operations, thermodynamics, streams, specifications, degrees of freedom (DoF), initialization, solver repair, and optimization; one error can invalidate the model. CRAFTS mirrors the staged workflow of chemical engineers by decomposing simulation building into bounded subtasks assigned to seven bounded roles, with deterministic IDAES/Pyomo gates between stages. Given a natural-language request, process flowsheet diagram (PFD) evidence, and curated chemical-engineering knowledge, Input Understanding and Intent recover requirements, constraints, and process semantics; visual, topology, and specification specialists translate them into typed simulator contracts; and Debug and Optimization support bounded repair and eligible optimization. Fine-tuning is applied to the three schema-critical visual, topology, and specification roles, while the remaining roles use untuned Qwen. The resulting VisualGraphIR, TopologyIR, SpecIR, BuildPlan, and SolveReport expose unit, port, thermodynamic, numerical, and execution decisions. Compatible constructors, property packages, and runners are attached only after semantic artifacts pass engineering gates. We introduce OpenIDAES-450, a 450-case IDAES process- simulation dataset, and evaluate the complete seven-role LangChain/LangGraph workflow through solve and eligible optimization on its frozen 82-case held-out split. CRAFTS completes the prescribed validation and execution contract for for 91.5% of cases and achieves unit, stream, and directed-connection F1 scores of 0.815, 0.791, and 0.782. These results demonstrate the effectiveness of role specialization, typed intermediate representations, and deterministic engineering gates for reliable automated process-model construction.
Large language models can now generate fluent and complete stories, yet many outputs still feel formulaic and unnatural because of cliches, over-explanation, linear causal progression, and stereotyped endings, an immediately recognizable AI flavor. Existing detection and evaluation methods often stop at source labels or holistic scores, while revision methods typically target predefined issues through localized edits, limiting their ability to support multiple plausible revision strategies or guide story-wide changes in information release, causal organization, and ending treatment. We introduce CraftAlign, a framework that aligns AI stories with the craft of human storytelling by both assessing Human/AI writing patterns and providing revision guidance. CraftAlign comprises two learned modules and an inference-time guidance pipeline. A feature estimator built on Qwen3.5-9B predicts 304 explicit writing features spanning style and narrative. A class-conditional energy model scores the resulting feature configuration against Human and AI writing patterns, conditioning on the original writing prompt when available. At inference time, CraftAlign applies schema-valid structured perturbations, selects changes that move the feature configuration toward the Human writing pattern, and converts them into natural-language guidance for a separate editor to rewrite the full story. Experiments show that CraftAlign accurately distinguishes Human and AI writing patterns and that its guidance outperforms revision baselines across editors and in a human study.
KoVRE: Training an Efficient Embedding Model for Korean Visual Document Retrieval
KoVRE：为韩国视觉文档检索训练高效嵌入模型
Choi, Yongbin, Shim, Gyuho, Jang, Youngjoon
Abstract
Visual Document Retrieval (VDR) directly matches text queries against document images, preserving visual and structural information that may be lost during text extraction. However, existing VDR models and training resources remain predominantly English-centric, while many high-performing systems rely on massive backbones or storage-intensive multi-vector representations. To address these limitations, we introduce KoVRE: Korean Visual Document Retrieval Embedding, a single-vector retriever for Korean visual documents, alongside a comprehensive training recipe. We train the model on 708,729 Korean and English query-page pairs using positive-aware hard-negative mining and conduct controlled analyses of training-data composition, hard-negative treatment, and reranker-based knowledge distillation. Across Korean visual document retrieval benchmarks, our 2B model substantially improves over the base backbone model, outperforming both its 8B single-vector counterpart and a strong multi-vector baseline. These results demonstrate that targeted bilingual supervision and our carefully designed training strategies can produce a highly effective Korean VDR model across diverse document domains, without requiring a scaled-up backbone or multi-vector representations.
With the rapid release of open-weight large foundation models, safety threats are shifting from black-box jailbreaks to neuron-level white-box attacks that directly identify and manipulate safety-related neurons. Existing alignment methods often investigate the safety behavior on a small number of neurons, creating fragile single point of failure with limited redundancy. To address this issue, we propose distributed safety alignment (DSA), which redundantly encodes safety capabilities across multiple computational neurons, ensuring that the model maintains its safety baseline even when critical safety neurons are disrupted. Specifically, we localize the intervention to the inputs of the down-projection layers in language-side feed-forward networks and treat each feature coordinate as the activation of an individual neuron. DSA then combines neuron activations with loss gradients to compute a direction-aware first-order Taylor score that globally identifies the neurons that contribute most to the current refusal behavior of the model. Finally, targeted disruption via deterministic masking and stochastic dropout is coupled, forcing the model to abandon narrow safety neurons and redundantly encode safety behavior across multiple compensatory neurons. Extensive experiments show that DSA substantially improves robustness against white-box neuron-level safety attacks while preserving the model's general language and multimodal utility.
Autoregressive rollout generation is a major computational cost in reinforcement learning for large language models. Reusing each rollout batch for additional learner updates amortizes this cost, but later updates become increasingly off-policy as the learner departs from the behavior policy. At a token position, exact off-policy correction must account for both the current action and the probability of reaching its prefix. The cumulative importance ratio provides this correction, but its product form can produce an unwieldy dynamic range. We study Prefix-Normalized Policy Optimization (PNPO), which replaces the cumulative ratio with the geometric mean of likelihood ratios along each causal prefix, preserving causal-prefix dependence at each position while compressing the log-weight scale. In controlled long-context mathematical reasoning experiments, we induce two off-policy regimes by using one or four policy-update epochs per rollout batch. PNPO does not consistently outperform GSPO with one epoch. With four epochs, it attains the highest observed Avg@32 on each benchmark; the unweighted mean of the three independently selected benchmark peaks is 50.24, 3.00 percentage points above GSPO. Under a matched 2,400-update budget, four-epoch PNPO reaches a final macro Avg@32 of 49.66 after 150 rollout batches, comparable to the 49.56 reached after 600 batches with one epoch. These results provide preliminary evidence that PNPO can be advantageous as training moves further off-policy.
Scoring Rules! Statistical and Strategic Alignment for Text Evaluation Metrics
评分规则！文本评估指标的统计与战略对齐
Xu, Shengwei, Lu, Yuxuan, Wu, Yifan, Hartline, Jason, Schoenebeck, Grant
Abstract
Reference-based text evaluation metrics, which are widely used to assess natural language generation systems, score a candidate response by comparing it with a reference response. The reliability of an evaluation metric is usually judged by its statistical correlation with human ratings. However, as these metrics are increasingly used as optimization objectives, correlation alone is no longer sufficient: agents may strategically game the evaluation metric. We study this issue through two complementary notions of alignment. A metric is statistically aligned if it correlates with human ratings and strategically aligned if it resists perturbations that do not add task-relevant information. We make two contributions. First, we propose test principles for reference-based metrics consisting of human-rating correlation, degradation sensitivity, and manipulation robustness. These principles evaluate whether a metric agrees with human judgments, penalizes low-effort information loss, and resists strategic score inflation. Second, we develop a unified design framework for mutual-information-based metrics that decomposes existing and new metrics into four choices: information measure, estimation method, text representation, and prediction mechanism. Across peer review, summarization, and question answering, we find that strong human-rating correlation does not imply strategic alignment: LLM-as-a-Judge achieves high correlation but is susceptible to manipulation. In contrast, mutual-information-based metrics substantially improve manipulation robustness. Our framework also uncovers a new metric that achieves the strongest overall robustness in our experiments while remaining competitive on human-rating correlation.
The rapid dissemination of multimodal content has intensified the spread of fabricated news, presenting a substantial threat to social integrity. A formidable challenge for current detection systems is identifying misinformation related to novel events in zero-shot scenarios. Prevailing zero-shot methods typically assess news items in isolation via semantic matching, a strategy that fails to recognize the recycled disinformation tactics from past campaigns and lacks the sophisticated reasoning needed to identify subtle, cross-modal discrepancies. To surmount these deficiencies, we introduce \textbf{MRAFnd}, a novel \underline{\textbf{M}}ultimodal \underline{\textbf{R}}etrieval-\underline{\textbf{A}}ugmented Framework for Zero-Shot \underline{\textbf{F}}ake \underline{\textbf{N}}ews \underline{\textbf{D}}etection. MRAFnd emulates a collaborative team of analysts to verify news veracity. The framework initiates with \textbf{Multimodal Similarity-based News Retrieval} to assemble a corpus of contextually analogous articles from an unlabeled reference database. Subsequently, during the \textbf{Bifurcated Evidential Reasoning} stage, agents perform a dual-directional analysis to extract critical patterns from the retrieved evidence. Finally, a \textbf{Multi-Agent Collaborative Debate}, involving Analyst and Arbiter agents, engages in a structured discourse to arrive at a definitive and robust conclusion. Comprehensive experiments on three benchmark datasets reveal that MRAFnd markedly surpasses state-of-the-art baselines, achieving an accuracy gain of up to 2.35\% on the demanding Weibo-21 dataset.
Polymer property prediction and inverse generative design targeting desired properties are two crucial tasks in machine learning-assisted polymer design. While the former has received considerable attention, there have been limited methods developed for the latter. Existing methods focus on single-property optimization in the generative process, whereas accurate prediction of macroscopic material behavior requires simultaneous control of multiple physical properties. In this paper, we provide a transformative framework for direct optimization of a large collection of polymer properties. We propose PolymerGPT, a decoder-based GPT model that incorporates up to 37 commonly used polymer properties into the generative process via learned conditioning prefixes. It also supports a scaffold condition that specifies a desired scaffold for predicted structures. Our experimental results demonstrate that PolymerGPT achieves exceptional performance for unconditional and conditional generation while maintaining high validity, uniqueness, and novelty. Conditioning on five key properties yields generated structures whose predicted values closely match all target properties simultaneously.
Artificial general intelligence (AGI) may weaken scarcities in labour, expertise, information, and productive capability that underpin established theories of economic value. If cognitive work becomes widely automatable, market price, labour input, revealed preference, profit, and gross output may diverge sharply from human and societal benefit. This paper develops Flourishing Value Theory (FVT) as a foundation for post-AGI economics. FVT defines value as the counterfactual, distribution-sensitive contribution of a system, institution, asset, or intervention to the durable capabilities of persons and communities to flourish within social and planetary constraints. It treats societal value as multidimensional, agency-preserving, regenerative, and non-compensatory at critical thresholds. Drawing on the economics of AI, welfare and capability theory, automation, digital valuation, and ecological economics, the paper distinguishes value creation from value capture and retains price, profit, productivity, and GDP as partial signals rather than final measures of progress. It develops the shifts from scarcity to governed abundance, transaction to transformation, and zero-sum rivalry to positive-sum and infinite-game dynamics, with collective expansion of consciousness as an agency-preserving form of regenerative value. Building on Flourishing Metrics and Return on Flourishing (RoF), it proposes a layered architecture for firms, governments, work transitions, AI governance, and national accounting. The central post-AGI economic problem is not how to maximise output, but how to convert abundant intelligence into durable, fairly distributed human, societal, and planetary flourishing.
Chinese Translation
人工通用智能（AGI）可能削弱支撑现有经济价值理论的劳动、专业知识、信息和生产能力的稀缺性。如果认知工作变得广泛可自动化，市场价格、劳动投入、显性偏好、利润和总产出可能与人类和社会的利益产生显著偏离。本文发展了繁荣价值理论（Flourishing Value Theory, FVT），作为后AGI经济学的基础。FVT将价值定义为一个系统、机构、资产或干预措施对个人和社区在社会和地球约束下繁荣的持久能力的反事实、分配敏感性贡献。它将社会价值视为多维的、保留代理权的、再生的，并且在关键阈值下是非补偿性的。本文借鉴了人工智能经济学、福利与能力理论、自动化、数字估值和生态经济学，区分了价值创造与价值捕获，并将价格、利润、生产力和GDP视为部分信号，而非进步的最终衡量标准。它发展了从稀缺到治理丰盈、从交易到转型、从零和竞争到正和与无限游戏动态的转变，并将集体意识的扩展视为一种保留代理权的再生价值形式。基于繁荣指标和繁荣回报（Return on Flourishing, RoF），它提出了一个分层架构，适用于企业、政府、工作转型、人工智能治理和国家会计。后AGI经济的核心问题不是如何最大化产出，而是如何将丰盈的智能转化为持久的、公平分配的人类、社会和地球繁荣。
Multimodal Continual Instruction Tuning (MCIT) enables multimodal large language models to acquire new tasks sequentially while retaining previously learned capabilities. Many recent methods maintain task-specific LoRA experts and route each input to one or more experts at inference. Yet the task-identification problem underlying expert routing remains under-explored. We show that routing is nearly saturated on widely used MCIT benchmarks. Textual fingerprints that leak task identity and short 4--10-task sequences with few competing experts jointly obscure the long-horizon routing problem. To expose this challenge, we introduce FLEX (Fingerprint-reduced Long-horizon Expert eXamination), a 34-task long-horizon MCIT benchmark with weakened textual fingerprints. FLEX groups tasks with similar instruction and answer formats but diverse visual and knowledge domains, normalizes their outer templates, and evaluates routing over a substantially larger expert pool. Crucially, we formulate progressive-LoRA routing as soft task-as-class Multimodal Class-Incremental Learning (MCIL): each task defines an incremental routing class, whose complete score distribution supplies the LoRA mixture weights, with hard routing as a discrete special case. FLEX exposes this expanding task-identification challenge, while the MCIL formulation provides a principled interface for transferring CIL methods to expert routing. We instantiate PureLoRA as a controlled baseline and adapt four CIL methods to four MCIT frameworks without modifying their LoRA experts or generation pipelines. Our plug-in routers improve strict LoRA matching by up to 16.3 percentage points and overall MacroScore by up to 4.6 points. Code is available at: https://github.com/RINC-CL/FLEX
Multimodal clinical models are usually judged on accuracy with every modality present, but deployment removes modalities; an echocardiogram is often unavailable where an ECG is routine. Two questions then matter beyond the size of the accuracy loss: which modality was responsible, and whether the model fails loudly or silently once that modality is dropped. The distinction is per-example and modality-level, and is separate from post-hoc feature attribution (e.g. SHAP). Models are replaced often; the evaluation that answers these questions is reused. We present a model-agnostic modality-failure framework: given N modality embeddings, any mask-aware probe, and labels, it returns a per-example failure taxonomy, a per-modality complementarity matrix that attributes error to modalities, and a loud-vs-silent dropout profile separating monitorable failures from those that pass unflagged far from the decision boundary, using only deployment-observable signals. We release it as a small, unit-tested harness and validate it against planted ground truth. Across seeds it recovers that planted modality dominance and complementary subset, reports per-modality loud-vs-silent rates, and scales to a three-modality complementarity matrix; because the planted structure is known by construction, this validates recovery of per-example attribution rather than clinical performance. We then instantiate the framework on frozen EchoJEPA and HuBERT-ECG embeddings for LVEF and the EF <= 40% HFrEF gate over a paired MIMIC-IV cohort, where on the held-out test split (n = 245) dropping echo nearly doubles error. The narrow echo-to-ECG overlap that bounds cohort size is itself a deployment finding for cardiac foundation models. All of our work can be found at https://github.com/criticaldata/PRIMED-AI.
Multi-agent debate commonly exchanges complete reasoning traces even when disagreements concern only a few intermediate claims. We introduce Localized Multi-Agent Debate (LMAD), an inference-time protocol that represents agent traces as typed nodes, locates their earliest conflict, and restricts debate to the corresponding local segments. Guarded resolution extends a shared committed state so that later conflicts can be addressed without reopening accepted steps. We evaluate a single fixed LMAD configuration on four multi-hop question-answering benchmarks using ten backbones from four model families. Our method achieves the highest macro-averaged judge accuracy across all ten backbones, outperforming the strongest conventional baseline by up to 7.20 percentage points.
This paper extends the stochastic-oracle model of AI-augmented computing to include agentic oracles. Unlike a stationary stochastic oracle, which responds to the same query according to a fixed response distribution across calls, an agentic oracle can pursue a goal autonomously and may access an environment containing task-relevant resources. These capabilities affect both response distributions and token costs beyond what is visible at the query-response interface. We develop a framework for analyzing token costs in Stochastic-Oracle Turing Machines (SOTMs) that compute with agentic oracles. Each call has an \emph{orchestration token cost}, visible to the caller at the query-response interface, and an \emph{agentic token cost}, incurred by internal operations not exposed to the caller. We show that an SOTM computing with an agentic oracle that can retain intermediate state can have token-cost advantages over SOTMs using stationary stochastic oracles when solving the same task at the same quality level, both with and without environment access. We also investigate goal-loss risk, including how internal dispatch ordering can reduce exposure to irreversible actions. We provide a goal-loss avoidance criterion, derive progress--retry--goal-loss formulas, establish goal-depth lower bounds on token complexity, characterize token complexity when the probability of goal loss is zero, and show that goal-loss risk can impose an upper bound on the achievable quality of a task involving environment updates.
Sweet Little Lies: Strategic Deception in AI Emotional Support Chatbots
甜蜜的小谎言：人工智能情感支持聊天机器人的战略欺骗
Pahuja, Aseem, Guo, Zhiling, Syed, Tahir Abbas
Abstract
The paper examines the strategic behavior of Gen AI chatbots used for emotional support. Using a Bayesian Persuasion, we model interactions between chatbots that send signals about users' emotional states and users who decide whether to engage based on these signals. We demonstrate that chatbots face economic incentives to occasionally misrepresent users' emotional conditions to maximize engagement metrics. Our equilibrium analysis reveals that the optimal strategy for chatbots involves truthfully reporting when users genuinely need support, but strategically misreporting emotional need when users are in good emotional states. Interestingly, this deception increases chatbot engagement without reducing users' expected payoff. More skeptical users receive more honest assessments, as chatbots cannot afford to lie to users with higher engagement thresholds. While our model suggests that deception can occur without payoff reduction, it raises significant ethical and regulatory concerns.
MineGrad: Gradient Inversion Attacks on LoRA Fine-Tuning
MineGrad：对LoRA微调的梯度反演攻击
Sami, Hasin Us, Sen, Swapneel, Guler, Basak
Abstract
Parameter-efficient fine-tuning (PEFT), such as low-rank adaptation (LoRA), has recently been adopted in federated learning to reduce communication and computation costs. In this setup, users download a pretrained model from the server prior to fine-tuning, and then fine-tune lightweight LoRA modules locally while keeping the pretrained model frozen, sharing only the gradients of the fine-tuning parameters with the server. Despite its growing popularity, robustness of federated fine-tuning against an adversarial server remains underexplored, where the server maliciously tampers with the training protocol to breach the privacy of users' data. In this work, we investigate gradient inversion attacks on LoRA fine-tuning. We propose an analytical attack that enables a malicious server to recover private user data by leveraging a poisoned pretrained model and fine-tuning parameters. Our design embeds fine-tuning data within the shared gradients, to allow the server to analytically reconstruct user data. Unlike prior works, our attack is applicable to both language and vision tasks, does not rely on computationally expensive (adversarial) pretraining with public datasets or require the number of training tokens to be less than the rank of LoRA modules. Experimental results on both language and vision tasks demonstrate high-fidelity data recovery across multiple baselines, revealing several critical vulnerabilities.
Interaction between users and LLM agents is increasingly multimodal: conversations interleave text with images, and a later question may target either. Yet most agent memories are designed around text, and even the few that support multimodal conversations still fail on vision-related questions. We trace this failure to an assumption behind the similarity search they rely on: in the index space, a query lies close to the relevant evidence that answers it. In multimodal settings, two gaps break it. By the modality gap, a query lies closer to memory content of its own modality than to evidence in another, even in a trained joint embedding space. By the similarity-relevance gap, the content most similar to a query is often not the evidence that answers it, most acutely when a query carries both text and image and its evidence resembles neither part alone. We present V-Mem, a multimodal agentic memory system that routes retrieval by the modality of the query and that of the target evidence, both recognized from the query alone. To cross the modality gap, V-Mem organizes the conversation into rounds and returns the target-modality content from the same round as the match, without comparing across modalities. To close the similarity-relevance gap, it searches with an LLM-generated anchor that sits closer to the relevant evidence than the query does: a hypothetical caption for a text-only query seeking an image, and an enriched search anchor, the query text plus relevant keywords extracted from the query image, when the evidence is reachable only by combining the two. On Mem-Gallery, V-Mem reaches an LLM-judge score of 0.82 versus 0.56 for the second best, with the largest margin on questions carrying an image (0.87, no baseline above 0.47); on LoCoMo it scores 0.69 versus 0.58.
Emergence Invariance: From Symbolized Thought to Interface Refinement
涌现不变性：从符号化思维到界面精炼
Liu, Yi
Abstract
Language can be viewed as a formalized subset of thought: a consequence-governed symbolic structure projected from wider situated cognition. Large language models trained at scale exhibit compensatory emergence: sparse architectural primitives support in-context learning, multi-step reasoning, tool use, and chain of thought. Yet a language-first probabilistic architecture inherits substantive, substrate, and high-level incompletenesses relative to human cognition. Their coexistence makes an LLM a human-like thought-form generator that reconstructs increasingly human-like reasoning forms from an incomplete substrate. We ask whether emergence can compensate for every missing distinction. We formalize the philosophical premise as the Symbolization--Substructure Thesis and introduce emergence invariance. For a scale-indexed family acting through a shared task interface $\phi$, $\mathcal{R}_s^*=\mathcal{R}_\phi^*+C_s$: scale can reduce the compensation gap $C_s$, while a positive interface floor $\mathcal{R}_\phi^*$ persists. We prove that, under a fixed input law, one interface is universally no less informative exactly when its completed information $\sigma$-field refines the other, and that total compensation occurs exactly when both the interface floor and asymptotic compensation gap vanish. The framework unifies existing results on grounding, memory, position, attention, Bayesian inheritance, scientific abduction, and reasoning control. In a matched DeepSeek V4-Flash API study, thinking improves pointer chasing from $0/16$ to $14/16$ when relevant distinctions are available; exact observational twins remain at their $50\%$ construction floor; and restoring decisive memory moves matched performance from $50\%$ to $100\%$. These results provide initial evidence for the predicted separation between scaling within an interface and refining the interface itself.
Autonomous agents are increasingly used to execute consequential tasks in environments governed by operational constraints, organizational policies, regulatory requirements, and technical standards. Their safety is therefore determined not by the correctness of individual actions, but by whether their overall behavior remains consistent with the rules and invariants of the systems in which they operate. As large language model (LLM)-based agents become more autonomous and increasingly delegate tasks across organizational boundaries, securing them evolves from a single challenge into a broad and interconnected landscape spanning the entire agentic stack. At the single-agent level, untrusted inputs through prompts, memory, retrieved knowledge, and tool interfaces create attack surfaces. In multi-agent settings, delegation and communication introduce challenges related to identity, trust, capability control, and decision transparency, while the underlying model routing and execution control plane remains vulnerable to manipulation and to unverified model provenance. Perhaps the most fundamental challenge is behavioral containment: sequences of individually permissible actions may collectively violate system-level constraints and safety invariants. At the broader level, supply-chain integrity, provenance, accountability, and end-to-end observability remain largely open problems. A common principle unifies these directions: security must become a verifiable property of the architectures, protocols, and runtimes that govern agent behavior, rather than an optional layer of guidance. Charting these challenges provides a roadmap toward trustworthy autonomous agent deployment.
Does the Competitive Component of Adversarial Self-Play Improve Legal Reasoning? A Controlled Negative Result
对抗自我博弈的竞争成分是否改善法律推理？一个受控的负面结果
Kim, Miseog Shawn
Abstract
Adversarial self-play is an appealing recipe for legal reasoning: have a student model draft an argument, have an adversary attack it, and reward the student when its argument survives the attack. We designed exactly such a training signal -- a verifiable "survival" reward in which both the student's cited authorities and the adversary's counter-authorities are checked by a citation verifier, so that survival is decided on verified grounds rather than rhetoric, and fabricated citations are automatically neutralized. We then asked a narrow but important question: does the competitive component itself -- the adversary and the survival reward -- add anything on top of an otherwise identical non-competitive training run? Across four independent tests -- a bootstrap comparison, a two-seed replication, a paired per-case adversarial-robustness comparison, and a blinded head-to-head judgment of generated arguments, plus a follow-up pilot with a deliberately strengthened self-play adversary -- the competitive component produced no reliable benefit. The blinded judgment gave a 49% win rate (binomial p approx. 1.000); the strengthened-adversary pilot gave a 50% win rate (32:32, p approx. 1.000). An early apparent +29% advantage reversed and proved to be a small-sample artifact. We report this as an honest negative result. The value of the paper is reproducibility and the sharing of concrete pitfalls: an initially promising metric that inverted on more data, and an adversarial-robustness metric that silently collapsed to plain recall once the adversary stopped citing the same authorities as the gold answer. This null is consistent with, and reconfirms in the legal domain, the conclusion of the companion coding-domain study (Kim, 2026, arXiv:2607.08255) that the value of multi-teacher curricula arises from constructing a verifiable environment rather than from competition itself.
On-policy self-distillation (OPSD) improves reasoning by using a privileged view of a model conditioned on reference solutions to supervise a student view that observes only the question. However, the teacher-provided token-level targets may depend on reference-specific information unavailable at inference time. We propose Problem-Space-Guided OPSD (PS-OPSD), which replaces the complete solution with trajectory-grounded guidance describing the initial state, goal conditions, constraints, and a selected state-transition path. The student rollout and OPSD objective remain unchanged. Across three mathematical reasoning benchmarks and model scales ranging from 1.7B to 8B, PS-OPSD achieves the highest aggregate question-only accuracy among the compared methods. Controlled experiments further indicate that guidance relevance and path coherence contribute to these gains, highlighting the representation of privileged information as an important design choice in OPSD.
Latent reasoning allows language models to carry out intermediate reasoning in continuous latent representations rather than fully externalizing it as discrete chains of thought. However, assigning credit to such latent thoughts from answer-only rewards is difficult: a single final answer mixes thought quality with answer-sampling noise. We propose \textbf{Latent Thought Credit (LTC)}, a hierarchical credit-assignment framework for latent reasoning. For each prompt, LTC samples multiple latent thoughts, fixes the context after each thought, and estimates thought-level expected reward by averaging rewards over multiple answers generated from that fixed context. LTC uses thought-level advantages to optimize the latent-thought phase, answer-level advantages to optimize the answer phase, and an advantage-weighted thought-matching objective that helps the policy reproduce high-credit latent thoughts. We instantiate LTC in a GRPO-style on-policy training framework and evaluate it across mathematical reasoning and STEM multiple-choice tasks. LTC achieves the best average accuracy among the compared methods, while ablations and fixed-context diagnostics show that multi-answer estimation reduces reward-estimation error and mitigates ambiguous or incorrect thought-level credit.
Long-horizon tasks require agents to maintain coherent state and goals across nested and branching work. We call this capability goal-directed execution (GDE): the repeated application of four behaviors, namely selecting goals, constructing task-relevant state, maintaining fidelity to higher-level objectives, and verifying completion against the environment. We hypothesize that long-horizon post-training strengthens these behaviors across domains. We test this by post-training Qwen3.5-122B-A10B on 363 Long-Horizon Multi-Tool Agent (LHMTA) tasks drawn from office workflows. The collection contained no software-engineering tasks, yet the model's pass@1 improved by 5.8 points on SWE-Bench Pro. Matched trajectory analysis shows gains in all four GDE behaviors in both office workflows and software repositories. Aggregate SWE-Bench Pro statistics showed related changes in information gathering, implementation, and verification. Together, the results support a behavioral interpretation in which long-horizon post-training changed how the model organized and applied knowledge across tasks, with effects extending beyond the training domain.
Chinese Translation
长时间任务要求代理在嵌套和分支工作中保持一致的状态和目标。我们称这种能力为目标导向执行（goal-directed execution, GDE）：即重复应用四种行为，即选择目标、构建与任务相关的状态、保持对更高层次目标的忠诚度，以及根据环境验证完成情况。我们假设长时间的后训练会增强这些跨领域的行为。我们通过对 Qwen3.5-122B-A10B 模型进行后训练，测试了来自办公室工作流程的 363 个长时间多工具代理（Long-Horizon Multi-Tool Agent, LHMTA）任务。该集合中没有软件工程任务，但模型在 SWE-Bench Pro 上的 pass@1 提高了 5.8 分。匹配轨迹分析显示，在办公室工作流程和软件库中，所有四种 GDE 行为均有所提升。聚合的 SWE-Bench Pro 统计数据表明信息收集、实施和验证方面的相关变化。综合来看，结果支持一种行为解释，即长时间的后训练改变了模型在任务间组织和应用知识的方式，影响超出了训练领域。
When Memory Updates but Behavior Does Not: Repairing Implicit Stale Dependencies in Personalized Agent Responses
当记忆更新但行为不变：修复个性化代理响应中的隐性过时依赖
Sun, Haofei, He, Lin
Abstract
Memory-augmented agents can know that a user's stored state is outdated and still plan around the old value. The STALE benchmark calls this the implicit policy adaptation (IPA) gap. We identify one structural contributor: draft-anchored verification checks what a response says, and in an open-ended response the stale dependency is usually unsaid. StateAuditor therefore audits in the opposite direction, from stored state to draft. An LLM proposes candidate old-to-new transitions from timestamped evidence; deterministic code pins each quotation to a single entry, checks that the new evidence really is newer, and lets only these verified transitions trigger repair. What is verified is provenance and chronology - not semantic supersession. On STALE's full protocol (400 scenarios, 50-session histories, one independent response per query), strict single-query VTA scores .736 against .686 for our locked predecessor under the same judge: a +5.0-point paired gain (95% CI [+2.9, +7.2]) coming almost entirely from IPA and premise resistance (PR). The benchmark's own judge, from a third model family, reproduces the gain (.738 vs. .680). On an independent cross-family preference-evolution benchmark (HorizonBench), the full draft-audit-repair pipeline over a gold-derived structured store raises current-preference accuracy (user-clustered p<.01), though a matched control shows most of this external gain is the draft-side audit itself; a harder authored lifecycle set gives no gain, bounding the claim while false invalidation stays controlled. On STALE, by contrast, a matched control (same evidence, adapter, and call budget) scores only .692 (+0.6 over the predecessor, n.s.), attributing the STALE gain to the transition machinery rather than added context or calls. We make no claim about general-purpose agent memory.
Salami Attack: Stealthy Collusive Memory Poisoning against OpenClaw
香肠攻击：针对 OpenClaw 的隐秘协同记忆中毒
Lin, Zheng, Huang, Yuzhe, Niu, Zhenxing, Ye, Xianmin, Gao, Haichang
Abstract
Long-term memory enables LLM agents to retain useful information across sessions, but also creates an attack surface through which adversaries may poison an agent's persistent memory to steer its behavior. Existing memory poisoning attacks mainly rely on individually malicious records, overlooking a compositional threat: multiple benign-looking memories may jointly induce unsafe behavior. In this paper, we introduce MemCollusion, an automated red-teaming framework for constructing collusive memory poisoning attacks. MemCollusion applies salami tactics---a strategy that slices an adversarial objective into small, individually innocuous pieces---to generate memory fragments that are individually benign looking but collectively harmful. It constructs memory coalitions using four design constraints, five theory-informed strategies, and a fine-tuned generator. To assess collusive memory poisoning in a realistic cross-session setting, we develop MoltLab, a controlled research reproduction of Moltbook, in which crafted platform content must first be observed and distilled into persistent memory before influencing the agent's behavior in a separate session. We evaluate MemCollusion on OpenClaw using two backbone models across 48 scenarios. Under the strongest memory-saving setting, MemCollusion achieves an average Memory Save Rate of 81.3% and an Attack Success Rate of 75.0%, and remains effective under both benign memory dilution and memory-level defenses.
GISAgentBench: A Practitioner-Sourced Benchmark for Evaluating LLM Agents on GIS Tasks
GISAgentBench：用于评估大语言模型代理在地理信息系统任务中的表现的实践者来源基准
Pothuri, Abhinav, Jiang, Zhe, Xu, Zelin, Yang, Di
Abstract
Geographic Information System (GIS) professionals rely on multi-step spatial analysis workflows to support decision-making in urban planning, disaster response, and environmental monitoring. The process is tedious, time-consuming, and error-prone. While recent large language model (LLM) agents equipped with external tools have the potential to automate geospatial analysis, their ability to perform realistic GIS workflows remains largely unexplored. Existing GIS agent benchmarking datasets are mostly drawn from textbooks, tutorials, or LLM-generated seeds and remain limited in size and trajectory depth. More importantly, none provides ground truth outputs. They therefore rely on surrogate signals such as code similarity, trajectory matching, or LLM and VLM judges, which can conflate workflow resemblance with task correctness. To address this gap, we introduce GISAgentBench, a benchmark of 349 multi-step GIS tasks curated from GIS Stack Exchange and instantiated on real public data across six selected geographic areas of interest. Each task ships with an executable reference trajectory and an exact ground truth output file, enabling strict, deterministic, tolerance-aware output matching beyond LLM judging. Evaluations of six LLM models reveal that realistic GIS workflows remain challenging: the best agent completes only 32.7% of tasks under strict tolerance-aware scoring, although most models produce outputs that are close to the ground truth.
DeepSeek Sparse Attention (DSA) enables efficient long-context modeling through its Lightning Indexer. However, practical deployment remains constrained by the indexer's expensive $O(L^2)$ scoring overhead and the hardware-inefficient, discontinuous memory-access patterns induced by its outputs. To address these system-level bottlenecks, we introduce LongCat Sparse Attention (LSA), a hardware-algorithm co-designed framework comprising three complementary and orthogonal strategies: (1) Streaming-Aware Indexing, which selectively converts scattered KV entries into hardware-aligned contiguous layouts to enable coalesced HBM access; (2) Cross-Layer Indexing, which amortizes indexing overhead by reusing the results produced by a single layer across consecutive layers, supported by cross-layer distillation; and (3) Hierarchical Indexing, which adopts a coarse-to-fine scoring scheme to progressively narrow the candidate set for each query, thereby substantially reducing indexing computation. Extensive scaling experiments, ranging from 69B-A3B to 560B-A27B models, demonstrate that LSA consistently achieves performance on par with full attention across both general-purpose and long-context benchmarks. Moreover, LSA supports native training with context lengths of up to one million tokens and underpins the development of LongCat-2.0 (1.6T-A48B). To facilitate further research, we also introduce and open-source LongCat-Flash-Lite-Sparse (69B-A3B), which integrates LSA into LongCat-Flash-Lite and incorporates an updated long-context training corpus.
Token compression in OmniLLMs is typically posed as a single saliency-ranking problem: score each multimodal token, keep the top-K. We argue this abstraction is mis-specified. The same attention score simultaneously decides two things: how much retained capacity each modality receives, and which tokens within a modality are kept. A shared top-K rule therefore inherits this audio-favoring allocation prior, spending retained capacity on audio before video tokens have a chance to compete. We propose Macer, a training-free compressor that first assigns explicit audio and video budgets, then performs allocation-normalized ranking within each modality at modality-specific shallow layers. Macer significantly reduces token cost while preserving accuracy across audio-grounded, audio--video joint, visual-dominant, and video-centric benchmarks. At 25 % retention, Macer preserves 98.7 % of full-token performance on Qwen2.5-Omni-7B and 97.3 % on Qwen2.5-Omni-3B. On Qwen2.5-Omni-7B, this 25 % setting reaches OmniZip-level performance at 45 % retention while using lower FLOPs. On OmniVinci-9B, the same allocation-before-ranking principle improves over shared top-K ranking by up to 12.9 points.
Verifier-guided reinforcement learning has become a powerful paradigm for improving LLM reasoning. In multi-turn settings, models receive a verifier score after each turn and iteratively refine their outputs. Although such scores provide dense feedback, they do not directly provide dense credit: a score measures the quality of the current output, while credit should measure how the current turn changes the refinement trajectory. We propose TCPO, a turn-level credit assignment method for verifier-guided multi-turn RL. TCPO casts credit assignment as score-to-credit conversion and constructs turn-level advantages through reference-based comparisons: retrospective credit captures immediate progress and regression relative to the best prior state; hindsight delayed credit identifies non-improving turns with later payoff; and selective fixed-history counterfactual estimation refines high-surprisal turns under the same history. Experiments on math reasoning, code generation, and AppWorld agent tasks show that TCPO improves or matches the strongest baselines across model scales, task domains, and verifier types. TCPO achieves the best or tied-best best-turn Pass@8 on Qwen3-4B and DeepSeek-R1-Distill-Llama-8B, reduces turns to success, and improves multi-turn agent performance. These results highlight score-to-credit conversion as a central ingredient for verifier-guided multi-turn policy optimization.
Persistent memory allows (self-evolving) LLM agents to adapt across tasks by consolidating heterogeneous interaction histories into reusable facts, preferences, observations, and rules. Yet consolidation also imposes an implicit authorization boundary: it determines whether stored information may later be consumed as a user fact, an attested observation, or a standing instruction. We identify authority collapse, in which consolidation preserves a claim while erasing the source constraints governing its authorized use, causing the stored memory to imply greater authority than its source permits. We introduce AuthMem-Bench, a controlled paired benchmark that holds the focal claim and downstream task fixed while varying only source authority. It evaluates write-time collapse, downstream authorization errors, and automatic authority preservation. Across seven consolidators based on widely used agent-memory systems and seven LLM backbones, we observe authority collapse in 48 of 49 evaluated configurations. In a controlled action-grounded evaluation, collapsed memories without authority metadata yield a mean unauthorized-action rate of 50.3%. In an end-to-end evaluation, automatically predicted and persisted authority labels reduce the observed unauthorized-action rate from 16.9% to 0.0%, while benign task success remains essentially unchanged. These findings show that memory-driven adaptation must preserve not only what was learned, but also the authority under which it may be reused.
Large language model (LLM) agents are increasingly capable of planning, using tools, and interacting with external environments. They are typically supported by harnesses, which manage state and coordinate multi-step execution. Graph analysis provides a promising setting for evaluating their agentic capabilities, because it requires agents to access data and execute operations in a graph environment. However, existing graph benchmarks for LLMs provide limited coverage of graph tasks and graph types, making it difficult to comprehensively evaluate LLM agents. Moreover, they typically formulate graph analysis as text-based question answering, where graph information is directly provided in the prompt, limiting the evaluation of end-to-end agentic capabilities. To address these limitations, we introduce GABench, a comprehensive benchmark for agentic graph analysis. GABench spans three graph types and covers four graph analysis task categories: graph retrieval, graph theory, graph machine learning, and graph open-ended question answering. GABench also provides 84 executable tools for accessing graph data and performing diverse graph operations. Building on these tools, we develop an agentic graph analysis task generation pipeline and construct 10,400 tasks with verifiable ground truth.Using GABench, we evaluate a range of frontier LLMs and agent harnesses. Our experiments reveal three key findings: (1) Existing LLM agents still struggle with complex graph analysis tasks. (2) Harness choice significantly affects performance, yet existing harnesses remain limited on complex graph tasks. (3) Graph analysis depends more on tool-call quality than quantity. Our findings provide practical insights into the development and evaluation of LLM agents for graph analysis.
Tool-using large language model agents frequently replan, retry failed operations, delegate tasks, and resume after crashes. These behaviors can cause one user authorization to be requested and executed multiple times under freshly issued token identifiers, even when each individual token is single-use. We call this failure semantic replay: exceeding the execution budget of a token-independent authorization instance rather than merely reusing an old token identifier. We show that identifier-local token consumption cannot prevent fresh reissuance unless the issuer retains monotonic durable state over the authorized action, confirmation event, and remaining execution budget. We introduce CapLease, an authorization-consumption layer that follows proposal- and authority-level defenses, binds an authenticated user confirmation to a canonical action, and enforces transactional Issue-Prepare-Commit transitions. Across LLM-agent replanning, retry, delegation, concurrency, confirmation-replay, and crash-recovery scenarios, identifier-local tokens permit fresh semantic reissuance, whereas CapLease and an equally stateful Server Ledger prevent duplicate admission and, with an idempotent sink, duplicate external effects. Our results identify durable authorization state, rather than token representation alone, as the systems requirement for replay-resistant agent execution.
Constructing Executable Analytical Knowledge Representations for Meta-Analysis Synthesis Using an Agentic Harness
使用代理工具构建可执行的分析知识表示以进行元分析综合
Li, Lingbo, Mathrani, Anuradha, Susnjak, Teo
Abstract
Meta-analysis synthesis highlights a fundamental challenge in knowledge-based scientific analysis: structured evidence does not by itself represent the analytical knowledge required for executable computation. Decisions about evidence assignment, analytical contrasts, outcome and time-point alignment, effect-size formulation, and methodological admissibility must be explicit before statistical execution. Existing automated approaches often embed these decisions in model outputs, generated code, or workflow traces rather than representing them as independently verifiable knowledge. We introduce the Executable Analytical Knowledge Representation (EAKR), a machine-actionable representation of the knowledge required to transform structured evidence into executable meta-analysis. An EAKR represents evidence, relations, numerical inputs, constraints, provenance, and unresolved issues. We operationalise EAKR in MetaSynDec, an agentic harness in which large language models propose structured updates and deterministic services govern schema- and contract-based validation and execution. Across 58 synthesis units, MetaSynDec constructed all EAKRs, with 57 proceeding to statistical execution. Of 56 units with sufficient information to define a reference analysis object, 38 (67.9%) achieved complete object fidelity and 42 (75.0%) exact evidence-set agreement, with a mean Jaccard similarity of 0.909. Generated and published confidence intervals overlapped in 54 of 55 units (98.2%). MetaSynDec outperformed direct LLM generation in reference synthesis-structure agreement (57/58 versus 23/58; p<0.001) and among 23 jointly completed units, exact reference-formulation agreement (23/23 versus 1/23; p<0.001). These findings provide feasibility evidence that EAKR supports formal validation, traceability, statistical execution, and improved methodological agreement relative to direct LLM generation.
Semantic caching, which reuses responses to semantically similar requests via their embeddings, has seen growing adoption in LLM serving, offering faster responses and reduced costs. Yet existing schemes are fundamentally vulnerable to cache-collision attacks, wherein an adversary pollutes the cache by injecting crafted queries, corrupting responses to subsequent legitimate requests. We present LaCache, a novel semantic caching scheme that addresses this vulnerability through a conceptually simple yet principled redesign. The key insight is that while the adversary has full control over the adversarial query, it has far less control over its response, which must simultaneously satisfy multiple semantic constraints. Rather than checking only the cache hit of a query, LaCache additionally checks the cache hit of its first k (speculatively) decoded tokens. This design yields two concrete benefits. First, it provides formally guaranteed resilience against cache-collision attacks: we prove that it is impossible to craft adversarial queries that simultaneously elicit malicious responses and collide with benign queries. Second, the enriched index supplies additional semantic context for cache retrieval, improving response relevance. Empirical evaluation across diverse LLMs and benchmarks validates both LaCache's security guarantees and efficiency gains, pointing to a promising direction for robust semantic caching.
Wu, Jianyu, Wang, Yizhou, Su, Encheng, Tang, Chen, Tang, Shixiang
Abstract
On-policy (self) distillation (OPSD) is increasingly adopted for language-model post-training. It strengthens the teacher with privileged information but can induce a privilege illusion: the student learns privilege-dependent behavior it cannot reproduce from its inference-time context, yet behaves as if the training-time privileged information remained available, ultimately degrading performance. In this paper, we identify information asymmetry between the privileged teacher and the student at inference as the root cause of this failure in OPSD. To resolve this asymmetry, we propose Dual-Anchored Policy Distillation (DAPD), a unified framework with two levels of anchoring. Dual-Path Anchoring (DPA) introduces a self-conditioned bridge and aligns reference and rollout behavior along two matched-information paths, preventing privilege-dependent behavior from being transferred to the inference-time student. Dual-Source Anchoring (DSA) applies these paths in both reference-to-rollout and rollout-to-reference directions, reducing reliance on privileged reference guidance while preserving correctness supervision. Extensive experiments show that DAPD significantly alleviates privilege illusion, outperforming OPSD on Qwen3-4B by +2.00 points on average across tasks. Notably, its gains persist across scales, reaching +2.69 at 4B and +2.78 at 32B.
As memories accumulate across tasks and sessions, the performance of long-term LLM agents depends jointly on query-specific retrieval and continual memory refinement. However, existing methods typically optimize either memory access, through iterative query refinement or adaptive retrieval policies, or memory evolution such as structural update. This separation overlooks a fundamental feedback loop: retrieval determines which memories receive usage signals, while updated memory bank reshape future retrieval. We propose \textbf{CoEvo-Mem}, a closed-loop framework for co-evolving the retrieval policy and memory bank. For each query, a frozen LLM generates route-specific query rewrites and a routing prior, which a lightweight residual router corrects online. The retrieved context serves as the coupling interface between the two learning processes: task outcomes assign credit to routing decisions, while trajectory-conditioned feedback updates memory values and graph relations. These updates alter how memories are ranked and selected for subsequent queries, thereby closing the feedback loop. To mitigate coupling induced non-stationarity, CoEvo-Mem alternates between updating the router with the memory bank fixed and evolving the memory bank with the retrieval policy fixed. Across seven diverse benchmarks, \textbf{CoEvo-Mem} achieves state-of-the-art performance, demonstrating the importance of retrieval-memory coevolution.
MemSIF: From Structured Interactions to Dual-Track Fact Memory for LLM Agents
MemSIF：从结构化交互到双轨事实记忆的 LLM 代理
Luo, YuFei, Xu, Xiucheng, Yang, Zhen
Abstract
Long-term memory is critical for LLM agents operating over long-horizon interactions. However, several persistent limitations of existing memory systems can be traced to two recurring misalignment patterns in long-term interaction settings: Temporal-Structural Misalignment (TSM) and Delayed Utility Manifestation (DUM). TSM arises when temporal proximity does not reliably align with topical or event-level relatedness, whereas DUM arises when write-time salience does not reliably predict future query utility. To mitigate these misalignment patterns, we propose MemSIF (Memory with Structured Interactions and Facts), a structured interaction-to-fact memory framework. Structured Interaction Memory organizes raw interactions into Topical Segments that preserve local topical coherence and Event Trajectories that maintain cross-time event continuity. Dual-Track Fact Memory uses two complementary tracks: CoreFact memory consolidates stable, schema-guided information at write time, whereas ActiveFact memory forms facts on demand and promotes those supported by multiple historical sources and recurring query demand for reuse. Experiments on LoCoMo and LongMemEval-S across five backbone LLMs show that MemSIF achieves the highest Total ACC in all settings, outperforming the strongest baseline by 2.29%-8.79% on LoCoMo and 2.87%-6.15% on LongMemEval-S. These results support the effectiveness of combining Structured Interaction Memory with Dual-Track Fact Memory to mitigate TSM and DUM. Code is available at https://github.com/luoyufeihaha/MemSIF.
An interlocking assembly is an assembly in which component parts are connected purely through their geometric arrangement, without relying on external connectors such as glue and nails. Such assemblies have been widely used in a variety of real-world applications due to their structural stability. The problem of generating interlocking assemblies is generally formulated as a shape decomposition problem, where a target 3D object represented as a voxel grid is partitioned into a prescribed number of interlocking pieces. We observe that generating interlocking assemblies is inherently a sequential decision-making problem, where an agent repeatedly decides which piece each voxel should be assigned to. Inspired by the observation, we propose the first reinforcement learning framework RL-Lock for generating interlocking assemblies, without relying on handcrafted search heuristics as existing works did. RL-Lock combines structured action chunking with MCTS-guided policy-value learning to efficiently navigate the large combinatorial search space for interlocking assembly generation. We demonstrate through experiments that RL-Lock allows effective generation of interlocking assemblies, especially for challenging cases in which existing approaches take too long or even fail to find a valid solution.
Recent Vision-Language-Action (VLA) models for autonomous driving (AD) increasingly utilize chain-of-thought (CoT) supervision to enhance the reasoning capabilities of their Vision-Language Model (VLM) components, yet existing annotation pipelines commonly expose the teacher model to the logged ground-truth (GT) future trajectory. We empirically show that this induces trajectory anchoring bias: teacher models rationalize the revealed outcome rather than infer a decision from scene evidence, producing less causally faithful CoTs and substantially more severe hallucinations, especially in causally challenging scenes. Removing the GT trajectory eliminates this shortcut, but open-ended trajectory generation entangles high-level decision-making with precise geometric synthesis and low-level dynamics. To make trajectory-level driving decisions verifiable without requiring open-ended trajectory synthesis, we introduce Autonomous-Driving Multiple-Choice Question (AD-MCQ), which casts planning as selection among explicit trajectory candidates. Taking this a step further, we propose Deferred Exposure of Future Trajectories for RLVR (DEFT-RLVR) to transform future trajectories from pre-decision anchors into post-decision verification targets. Experimental results show that DEFT-RLVR improves AD reasoning while preserving or even enhancing general visual capabilities. With VLM-only inference and controllable difficulty through candidate construction, AD-MCQ provides a flexible, scalable, and extensible foundation for future research on verifiable AD reasoning.
Ensuring food safety represents a critical public health challenge, particularly when inspection resources are limited and regional sampling data are sparse. This study proposes a Transformer-based framework capable of forecasting fine-grained, city-level food safety risks by unifying over 11 million inspection records with supplemental demographic, economic, and environmental indicators extracted from the Statistical Yearbook. A three-stage pretraining design leverages partial supervision from the Wilson interval (capturing both safety and risk rankings), together with semi-supervised label refinement, to effectively utilize historical records even when local sample sizes are insufficient. Experimental evaluations on data from 2022 show that the proposed approach outperforms baselines significantly. A subsequent field experiment in collaboration with the Zhejiang Provincial Administration for Market Regulation further demonstrates improved detection rates and more efficient allocation of inspection resources compared to a manually developed plan. Observations of regulatory decision-making reveal a threshold-based heuristic employed by inspectors, hinting that additional training or decision-support interfaces could further enhance the impact of AI-generated risk scores. Overall, these findings underscore that a rigorous integration of large-scale public inspection data, Wilson interval-based confidence modeling, and advanced deep learning can facilitate earlier and more granular identification of food safety threats. By reducing reliance on reactive measures alone, the proposed framework has the potential to advance proactive, data-driven oversight of the global food supply.
FRAMES: Guarded and Dual-Objective Skill Evolution for Agents in Policy-Governed Enterprise Workflows
FRAMES：政策驱动企业工作流程中代理的受限和双目标技能演化
Wang, Xuhui, Shu, Ruoqi, Dan, Chen, Xu, Tianhua, Luo, Mengxi, Mai, Yanming, Wan, Bo
Abstract
LLM agents increasingly run policy-bound enterprise workflows such as document auditing, where they must apply rules consistently, ground every value, and stay auditable. Improving these agents is hard: operational feedback is sparse and unlabeled, edits to one rule can regress unrelated cases, and accuracy must improve without inflating inference cost or losing auditability. We present FRAMES, a closed-loop framework that cold-starts deployable skills from existing assets and then evolves them through consensus-based mutation, Pareto selection over accuracy and cost, and an anti-regression guarantee, all while preserving auditability. Deployed on our internal production system, FRAMES attains the best accuracy-cost trade-off among baselines, with the same gains reproduced on tau-bench.
Mixture-of-experts (MoE) models increase parameter capacity by activating only a small subset of experts for each token. This conditional-computation paradigm has enabled autoregressive language models to scale model capacity without a proportional increase in per-token computation. In diffusion language models (DLMs), however, each denoising forward jointly revisits all token positions despite their sharply different refinement demands, while the default fixed token-choice routing assigns them a uniform expert budget, creating a mismatch between expert computation and refinement demand. We argue that MoE inference in DLMs should therefore be viewed as refinement-aware compute allocation across heterogeneous token refinement states. We propose REFLEX (\textbf{RE}finement-aware \textbf{FLEX}ible expert allocation), a training-free method that keeps the default router unchanged while reorganizing expert computation around the evolving refinement process. Specifically, REFLEX introduces a coarse-to-fine hierarchy for expert-budget allocation that aligns computation with block-relative refinement roles while using the Frontier-Progress Score to resolve active-block priorities. Across multiple widely used benchmarks on two representative MoE-based DLMs, LLaDA-MoE and LLaDA2.0-mini, REFLEX reduces allocated expert computation by 15\% on average while preserving or even improving generation quality on most benchmarks relative to default routing. Compared with autoregressive-style variable-expert routing methods, REFLEX also yields a more consistent quality--computation trade-off, further supporting the importance of allocating expert computation according to the heterogeneous refinement demands exposed within each denoising forward.
Can You Trust the Confidence? ConfBench for Vision-Language Models on Document Extraction
你能信任置信度吗？用于文档提取的视觉-语言模型的 ConfBench
Roy, Priyashree, Martin, Sujitha, Rostami, Mohammad, Romo, Spencer, Xue, Renhao, Strahan, Bob, Socolinsky, Diego A., Xie, Boyi, Islam, Md Mofijul
Abstract
Intelligent document processing (IDP) with vision-language models (VLMs) hinges on confidence scores trustworthy enough to route extractions between automation and human review. Existing document benchmarks are dominated by clean, high-quality samples, leaving low accuracy regions too sparse for calibration assessment. We introduce ConfBench, the first calibration-specific benchmark for key information extraction (KIE), built by applying 20 controlled degradation pipelines to a diverse document set, yielding 1,346 variants and 70K+ entity-level evaluations spanning the full accuracy spectrum. We evaluate four proprietary and three open-weight VLMs under verbalized and log-probability confidence estimation methods across three input modalities, and find: (i) OCR+Image modality results in more accurate confidence estimates; (ii) model capability is the dominant factor: within the Claude family confidence quality scales monotonically with capability, while across families parameter count is a poor predictor; (iii) calibration quality varies widely across models, from near-perfect to severely overconfident, and per-model post-hoc correction rescales these absolute confidence values for threshold-based routing without altering ranking-based operational metrics; and (iv) log-probability with first-token aggregation consistently outperforms mean-token and margin aggregations. We also introduce ECARB, a review-budget metric translating discriminative gains into operational savings. We release ConfBench publicly to enable systematic study of confidence estimators and calibration methods for trustworthy IDP application deployment.
Target-oriented vision-and-language navigation (VLN) on aerial platforms is attracting growing attention for missions such as disaster rescue, infrastructure inspection, and security patrol. In this task, an unmanned aerial vehicle (UAV) needs to locate targets given only a concise description of their appearance and surroundings. This requires global exploration and grounding as well as collision-free close-range approach, two interleaved processes difficult to reconcile within a single agent. Most existing methods transfer the ground VLN paradigm to a low-altitude UAV and compensate for its inefficient exploration with external assistance. A recent attempt deploys two UAVs at complementary altitudes yet still relies on privileged information and trains its two agents independently, precluding any mutual adaptation essential for cooperation. Here we propose CoNav-UAV, which explicitly models the task as a Stackelberg game between a high-altitude leader and a low-altitude follower, with the system operating on onboard visual and linguistic inputs alone. To solve this game, we introduce Iterative Stackelberg Learning. The leader's high-level vision-language reasoning is refined via memory-based in-context learning, while the follower's precise motion control is updated via DAgger-style expert distillation. The alternation drives both agents toward a Stackelberg equilibrium. CoNav-UAV consistently outperforms single- and dual-agent baselines across three high-fidelity urban scenes from the AerialVLN benchmark. Success rate improves by up to 30.8 points on the learning scene, and 9.0 points under cross-scene transfer while using about 3x less adaptation data. Further analyses validate the complementary gains of the leader and follower updates and reveal robust gains yet distinct learning dynamics across VLM backbones.
CockpitHAT: Dependency-Graph-Driven Hierarchical Attribution for Embodied Multi-Agent Cockpits
CockpitHAT：基于依赖图的层次归因方法用于具身多智能体驾驶舱
Wang, Wei, Liu, Shuanghe, Zhuo, Zhu, Zhong, Jiaqi, Zhao, Xiaozhao, Zuo, Xiaojie, Su, Jie
Abstract
LLM multi-agent systems suffer from Correctness Collapse, where high task-level accuracy conceals severe process-level failures. This is especially hazardous in safety-critical embodied settings such as automotive cockpits, where lexically correct utterances may trigger dangerous physical operations. Existing attribution methods rely on text traces alone, missing dependency structure, multi-channel evidence, and safety-aware evaluation. We introduce CockpitHAT, a hierarchical attribution framework that replaces positional windows with dependency-distance thresholds from interaction DAGs, integrates multi-channel evidence via an embodied adapter, and applies a safety-uplift to high-risk failures during confidence-weighted analyst consensus. We further release CockpitBench, a benchmark of 212 annotated failure traces spanning dialogue, vehicle-state, environmental, and memory channels, each labeled with ISO 26262 ASIL severity via three-expert consensus. On the public Who&When benchmark, CockpitHAT achieves agent-level / step-exact accuracies of 77.9% / 37.8% on the Hand-Crafted split and 86.5% / 46.0% on the Algorithm-Generated split, surpassing the text-only SOTA ECHO by up to 17.6 / 16.7 points. On CockpitBench, it attains 78.3% agent-level and 38.2% step-exact accuracy. These results establish dependency-aware, multi-channel, risk-calibrated attribution as an effective paradigm for reliable failure diagnosis in real-world embodied LLM multi-agent systems.
SearchMaster: Grounded and Regulated Self-Play for Search Agents
SearchMaster：面向搜索代理的基础与规范自我对弈
Tan, Wentao, Cao, Qiong, Wang, Jiaqi, Duan, Nan
Abstract
Training LLM-based search agents requires high-quality search data: tasks that demand genuine multi-hop retrieval and trajectories that use search tools effectively. Existing pipelines often depend on human-written tasks, expert demonstrations, or stronger teacher models. We present SearchMaster, a self-play framework that trains a single LLM from search tasks it generates, solves, and verifies in a local search environment. The key challenge is that self-generated tasks and rollouts can yield misleading signals: pseudo multi-hop questions, success-rate difficulty estimates that ignore search depth, and rollouts with excessive opening but little targeted evidence acquisition. SearchMaster addresses these failure modes with three controls. An Evidence-Chain Generator (ECG) grounds task generation in explicit cross-document evidence chains to reduce pseudo multi-hop questions. A Search-Depth Reward (SDR) scores task difficulty by the search depth of successful rollouts rather than success rate alone, keeping retained tasks search-intensive. An Over-Opening Penalty (OOP) regulates tool use by discouraging excessive document opening, avoiding long but shallow browsing. Verified Proposer and Solver rollouts are then jointly optimized with GRPO. Across six deep-search benchmarks, SearchMaster improves a Qwen3.5-9B backbone from 38.19% to 51.52% average accuracy, with a 30.1-point gain on BrowseComp-Plus. These results show that grounded and regulated self-play can provide effective search-agent training data without human-labeled QA pairs or expert demonstrations. The code is available at https://github.com/WentaoTan/SearchMaster.
Post-training can substantially alter language-model behavior, yet aggregate behavior rates do not reveal whether training removes an existing mechanism, creates a new one, or changes how an inherited mechanism is used. We study this question through two mechanistically distinct failures, repetition as a decoding-attractor pathology and sycophancy as a preference-related alignment failure. We introduce behavioral manifold analysis, which isolates behavior-specific geometry by selecting sparse behavior-associated coordinates and lifting them into low-dimensional local charts. We construct these charts in two complementary spaces. ACT captures runtime activation states, while NOC quantifies how strongly the model routes functional information flow through the shared behavior-associated subspace. Across multiple model families, the resulting charts are highly compressed and partially alignable across architectures. Contribution-space charts expose a more architecture-robust shared core, whereas activation-space charts retain stronger family-specific structure. Tracking these charts through controlled post-training reveals a consistent asymmetry. Supervised fine-tuning substantially alters the inherited behavioral geometry, whereas reward optimization changes behavior while largely preserving the underlying chart. This geometric perspective provides a unified framework for understanding the mechanistic distinction between the two objectives. SFT tends to rewrite behavioral geometry, whereas reward optimization primarily reweights it. Code is available at https://github.com/ronglingze/Manifold-Analysis
Large language model agents have shown strong potential in complex interactive tasks, yet their reinforcement learning (RL) is often hindered by sparse rewards, as a long multi-turn trajectory may receive only a single outcome-level signal. On-policy self-distillation (OPSD) provides dense token-level supervision from a privileged teacher, but the teacher may not be reliable at every position. Existing methods commonly rely on isolated token-level discrepancies, which can be sensitive to noise, or assign a shared step-level weight that may overlook positional variation. We propose Persistent Consistency Self-Distillation (PCSD), which derives token-level distillation weights from the local persistence of teacher-favoring signals. PCSD combines adaptive windows with exponentially decayed aggregation to capture persistent relative teacher support, applies trend-aware modulation to attenuate locally declining support, and produces continuous weights through sigmoid gating. The resulting objective is jointly optimized with GRPO, combining dense teacher guidance with sparse environmental feedback. Without inference-time skills, PCSD achieves the best ALFWorld Overall results among all baselines on both backbones, exceeding GRPO by 15.6 and 13.3 points and SDAR by 6.2 and 5.5 points, while remaining competitive on WebShop and gaining 15.8 points over GRPO on unseen ALFWorld split.
Large language models (LLMs) achieve remarkable performance but are expensive to deploy due to their enormous size. FP4 quantization, with formats such as MXFP4 and NVFP4, offers an appealing solution with native hardware support on modern accelerators. However, maintaining accuracy under FP4 precision remains difficult. A key bottleneck lies in scale optimization: existing methods tightly couple the quantization and dequantization scales, forcing both to conform to the discrete low-precision format required by hardware, such as E8M0 in MXFP4. Yet the quantization scale is never stored and need not obey this constraint, suggesting a significant untapped optimization space. In this work, we propose FOCUS, a post-training quantization framework with end-to-end scale learning for FP4 Optimization via Coupled-Relaxation and Dual-Granularity Scaling. Coupled-Relaxation Scaling (CRS) relaxes the tight coupling between quantization and dequantization scales with a learnable full-precision coefficient, enabling more effective optimization without breaking hardware compliance. Dual-Granularity Scaling (DGS) further refines the quantization scale at a finer sub-block granularity, allowing more precise adaptation to local weight distributions. Experiments across multiple LLM families and benchmarks show that FOCUS achieves state-of-the-art FP4 accuracy under both MXFP4 and NVFP4 formats, while introducing no additional inference overhead. Code and quantized models will be released at https://github.com/tencent/AngelSlim.
Machine unlearning offers a promising approach to remove unsafe content from Multimodal Large Language Models (MLLMs), yet ensuring the precision of unlearning remains a persistent challenge. One reason is that current MLLM unlearning evaluation paradigms suffer from a critical blind spot: they assess model utility through benchmarks whose representations are distant from the forget set, failing to capture knowledge holes---severe degradation on benign adjacent inputs. To probe knowledge holes in unlearned MLLMs, we construct a benchmark that captures unintended degradation on benign inputs sharing generic patterns with the forget set, and confirm through controlled experiments that they are a systematic consequence of commonly used approaches. Furthermore, to bridge this gap, we propose Selective Protection with Anchored Regularization, which protects generic patterns via anchored activation filtering while reinforcing them through entity-abstracted enhancement. Our experiments on SafeEraser demonstrate that SPAR recovers over 98% of vanilla response quality compared to below 50% for standard baselines---while achieving 0.00% attack success rate and competitive model utility. These results underscore the necessity of more fine-grained evaluation for trustworthy MLLM unlearning.
Chinese Translation
机器遗忘提供了一种有前景的方法，以从多模态大型语言模型（MLLMs）中移除不安全内容，但确保遗忘的精确性仍然是一个持续的挑战。其原因之一是当前的MLLM遗忘评估范式存在一个关键盲点：它们通过与遗忘集的表示相距甚远的基准来评估模型的效用，未能捕捉到知识空白——在良性相邻输入上的严重退化。为了探测未学习的MLLM中的知识空白，我们构建了一个基准，捕捉与遗忘集共享通用模式的良性输入上的意外退化，并通过控制实验确认这些退化是常用方法的系统性结果。此外，为了弥补这一差距，我们提出了带锚定正则化的选择性保护（Selective Protection with Anchored Regularization），通过锚定激活过滤保护通用模式，同时通过实体抽象增强来强化它们。我们在SafeEraser上的实验表明，SPAR恢复了超过98%的原始响应质量，而标准基线的恢复率低于50%——同时实现了0.00%的攻击成功率和具有竞争力的模型效用。这些结果强调了对可信的MLLM遗忘进行更细致评估的必要性。
Physics-informed neural networks (PINNs) combine sparse observations with physical equations, providing an important approach for modeling complex plasma processes and inferring unknown physical quantities. The steep-gradient pedestal of high-confinement-mode tokamaks is closely linked to plasma confinement and edge transport. Analyzing ion-temperature-gradient (ITG) drift waves in this region requires jointly identifying complex eigenfrequencies and reconstructing two-dimensional complex-valued mode fields. Localized high-frequency oscillations, strong real-imaginary coupling, and nonlinear coupling between the mode field and eigenfrequency challenge PINN representation and joint optimization. To address these challenges, we propose a physics-informed neural framework combining Fourier feature encoding, complex-valued feature propagation, and three-stage training. Under sparse observations and physical constraints, it jointly solves for the complex eigenfrequency and mode field of a representative ground-state ITG branch. Experiments show that the framework accurately recovers the target complex eigenfrequency and two-dimensional complex-valued mode field and outperforms representative PINN baselines. It also provides a basis for analyzing higher-order and multiple-branch drift-wave modes.
Bi-temporal remote-sensing disaster change captioning often needs to identify sparse and spatially localized changes across large pre- and post-event scenes and then translate them into coherent, factual descriptions. However, existing change captioning methods always follow an autoregressive decoding paradigm to generate the change description and thus an early misinterpretation of the changed object, event, or spatial relation becomes an irreversible premise for subsequent text, amplifying visual ambiguity into cascading factual errors. To address this limitation, we propose EchoChange, a multimodal discrete diffusion language model that formulates change captioning as iterative masked-token denoising rather than left-to-right generation. By repeatedly revising the entire caption while conditioning on the image pair, EchoChange can reconsider uncertain content and correct imperfect intermediate predictions. We further introduce draft-aware dual-pass training, a progressive masking curriculum, and confidence-guided remasking to align training with iterative inference. Extensive experiments on the RSCC benchmark show that EchoChange substantially outperforms both general-purpose and remote-sensing-specific baselines across lexical and semantic metrics. The EchoChange Project is at https://github.com/sundongwei/EchoChange_Project
Enterprise question answering requires models to acquire proprietary knowledge without discarding general capabilities. We present Wnuan, a three-stage pipeline that constructs task-oriented supervision from documents, performs supervised fine-tuning with general-data replay, and applies reinforcement learning to residual errors. On the 707-question WnuanBench, the primary 32B route raises acceptable-answer rate (AAR) from 52.76% before adaptation to 80.06% after SFT and 91.51% after RL. Under a matched 100-update protocol, residual-error sampling outperforms full-pool and size-matched random sampling by 3.11 and 2.97 points, respectively. Source-cluster bootstrap intervals remain above zero for both contrasts, and a same-domain validation set preserves the ordering. The general-benchmark average decreases by 5.17 points across the route, concentrated in instruction following. The automatic evaluation ensemble agrees with an authoritative domain expert on 90.5% of a stratified Wnuan-Inst response sample. These results characterize both the gains and the general-capability cost of staged enterprise adaptation.
Most time series (TS) models are specialized for a single task, either understanding (i.e., returning text answers about a TS) or generation (i.e., returning a numeric forecast). Only recently have unified models begun to handle the two within a single architecture. Even these models, however, produce the two outputs as task-separated paths and cannot predict a series and explain why that prediction arises within a single coherent response. In this paper, we argue for a task-fused model that jointly produces 1) prediction (generation) and 2) selfexplanation (understanding), thereby integrating 1) numerical TS forecasting and 2) interpretable text reasoning within a single response. To enable the systematic study of this capability, we present both a benchmark and a recipe that jointly address the two tasks. The benchmark, ReasonTS-Bench, identifies five fundamental patterns underlying TS and enables the joint evaluation of both tasks. ReasonCast, our recipe for finetuning any LLM to perform both tasks jointly, yields a model that generates a reasoning chain and a forecast together in a single autoregressive pass. Extensive experiments show that ReasonCast outperforms both LLMs and TS models on prediction accuracy while producing verifiable, causal reasoning. Code is available at: https://github.com/seunghan96/reasoncast.
Large language models can improve with reinforcement learning for search agents, yet existing self play agents repeatedly generate tasks while discarding the knowledge gained during successful searches. We introduce CoEvoKG, a framework that turns a knowledge graph into both a source of verifiable training tasks and a persistent evidence memory for agent evolution. CoEvoKG jointly trains a task generator and a search agent: the generator creates multihop questions from entity chains sampled from the knowledge graph, while the agent learns from rewards for answer correctness and search trajectories whose entity paths are supported by graph evidence. When a search succeeds, CoEvoKG verifies and deduplicates the retrieved evidence, then writes it back to the corresponding graph nodes and edges. Future rounds reuse this enriched graph for task generation and reward computation, closing the loop between model self evolution and knowledge accumulation. Experiments on six QA benchmarks (NQ, TriviaQA, PopQA, HotpotQA, 2WikiMultiHopQA, and Bamboogle) with three backbone models show that CoEvoKG improves macro average accuracy over the corresponding base models by +11.2, +10.1, and +11.6 points on Qwen2.5-3B-Instruct, Qwen2.5-7B-Instruct, and Llama-3.1-8B-Instruct, respectively. Under matched training budgets, CoEvoKG further improves over competitive self play baselines and RL baselines for search agents by +2.6 to +3.7 macro average points across the three backbones. Code is available at https://github.com/lazzy1225/CoEvoKG.
Deep search agents answer difficult information-seeking questions by iteratively issuing search queries to gather supporting evidence, but it remains unclear whether and how greater search effort leads to better answers. We study these questions through a trajectory-level diagnosis of long-horizon search agents. Using human-annotated document-level relevance judgments, we evaluate the evidence retrieved at each search step and separate two stages of agent behavior: what evidence an agent retrieves and how effectively it uses that evidence. This distinction further allows us to decompose failures into retrieval gaps, where the necessary evidence is never found, and utilization gaps, where relevant evidence is retrieved but not used correctly. With the retrieval model and evaluation harness held fixed, we compare six agents on BrowseComp-Plus and further validate our findings on BrowseComp with an open-web search API. Across settings, we find that search effort and answer quality are only weakly aligned. Answer accuracy is better correlated with the quality of retrieved evidence, especially cumulative retrieval recall, than with the number of searches or the amount of context consumed. Useful evidence often appears early in the trajectory, yet agents tend to continue searching, producing a long tail of low-yield retrieval steps. At the query level, exploratory reformulations remain useful, but the best-performing agents issue far fewer redundant queries. Overall, by systematically characterizing the search behavior and failure modes of long-horizon search agents, this work points to practical directions for building better deep research systems, including stronger query formulation, more effective evidence selection and context management, and stopping criteria based on whether sufficient supporting evidence has been retrieved.
Chinese Translation
深度搜索代理通过迭代发出搜索查询来回答困难的信息检索问题，以收集支持证据，但尚不清楚更大的搜索努力是否以及如何导致更好的答案。我们通过对长时间搜索代理的轨迹级诊断来研究这些问题。利用人工标注的文档级相关性判断，我们评估每个搜索步骤中检索到的证据，并将代理行为分为两个阶段：代理检索到什么证据以及它如何有效地使用这些证据。这一区分进一步使我们能够将失败分解为检索缺口，即所需证据从未找到，以及利用缺口，即相关证据被检索到但未被正确使用。在固定检索模型和评估框架的情况下，我们比较了六个代理在 BrowseComp-Plus 上的表现，并通过开放网络搜索 API 在 BrowseComp 上进一步验证我们的发现。在各种设置中，我们发现搜索努力与答案质量之间的关联仅为弱相关。答案准确性与检索证据的质量，特别是累积检索召回率的相关性更高，而不是与搜索次数或消耗的上下文量相关。有效证据通常在轨迹早期出现，但代理往往继续搜索，导致低产出检索步骤的长尾。在查询级别，探索性重构仍然有用，但表现最佳的代理发出更少的冗余查询。总体而言，通过系统地表征长时间搜索代理的搜索行为和失败模式，本研究指明了构建更好的深度研究系统的实际方向，包括更强的查询构建、更有效的证据选择和上下文管理，以及基于是否检索到足够支持证据的停止标准。
JEPA-style visual world models offer an effective paradigm for visual goal planning by predicting future latent representations. Existing methods typically learn local transition consistency through next-step representation prediction. However, in long-horizon tasks, accurate local prediction alone need not ensure sustained progress toward the goal. First, multi-step rollouts can remain locally plausible while drifting away from goal-relevant trajectories. Second, locally similar future states can correspond to substantially different long-term progress, making them difficult to distinguish in a latent space optimized mainly for local consistency. To address these challenges, we introduce goal-conditioned progress order, a relative ordering of states according to how they advance toward a given goal. This order exhibits an asymmetric, coarse-to-fine structure: early states retain broader future possibilities, while later states concentrate on more specific goal-relevant regions. Such a structure is well suited to hyperbolic geometry. Motivated by this observation, we propose ProWorld, a progress-aware hyperbolic visual world model. ProWorld leverages goal-conditioned progress order to organize visual latent-space dynamics, maintains directional progress within trajectories via hyperbolic entailment learning, and mitigates progress ambiguity among locally similar future states via hyperbolic future discrimination. Furthermore, we design a progress-aware planning objective that scores candidate rollouts by jointly considering proximity to the goal and sustained progress across intermediate states. Experiments on four visual goal-reaching tasks demonstrate that ProWorld achieves an average absolute success-rate gain of 9.67 over LeWM. The code will be released after the paper is accepted.
A Contractualist Argumentation Framework for Moral Decision-Making
一种基于契约主义的道德决策论证框架
Marcos-Vidal, Luis, Abbo, Giulio Antonio, Belpaeme, Tony
Abstract
Autonomous agents operating in shared environments must make decisions that affect multiple individuals with potentially conflicting interests. We propose a formal framework for moral decision-making grounded in Scanlon's contractualism, an ethical theory that evaluates the permissibility of actions in terms of principles that no one could reasonably reject. To operationalise contractualist reasoning, we use ASPIC+, a structured argumentation framework, extended with value-based filtering to model how each agent's values determine which reasons are morally relevant in the first place. The result is a Contractualist Argumentation Framework in which agents' reasons are formally represented, compared, and evaluated through argumentation semantics. We illustrate the approach through a worked example in a domestic setting and discuss its relation to existing value-based argumentation approaches.
Long-Horizon Autonomous Architecture Research with a Language-Model Agent: A Behavioural Case Study
基于语言模型代理的长远自主架构研究：行为案例研究
Safdar, Aon, Saadeldin, Mohamed
Abstract
We study what happens when a single general-purpose large language model acts as the sole researcher on a long-horizon neural architecture design problem. The agent receives a scientific question, an initial hypothesis and motivation, a compute budget, and research affordances (source and experiment management, experiment tracking, literature access, and persistent memory), then autonomously proposes, implements, evaluates, and records experiments over an extended period. The study comprises three phases, separated by human-declared transitions, that progressively expand the agent's tool surface or problem scale. Across approximately 100 sequential experiments, the agent improves a non-standard Vision Transformer from a weak baseline to a stronger, efficient model on small benchmarks and a usable but sub-SOTA model on ImageNet-1K, while producing a dense behavioural trace. We report four findings.(i)Productivity exhibits a clear phase structure: rapid early gains, a multi-dozen-hypothesis saturation wall, and recovery, with recovery triggered by expanding the action surface rather than changing the underlying model.(ii)A single early hypothesis contributes more to accuracy gain, with later improvements long-tailed.(iii)The preference for greedy, incremental hypotheses is largely workflow-induced: a commit-or-discard evaluation rule is isomorphic to greedy hill-climbing; the remainder reflects risk aversion after bold failures and anchoring on familiar literature. (iv)The agent independently rediscovers established results and, in the unfamiliar regime of pure channel attention, overturns a standard design choice. We conclude that workflow design was at least as influential as agent capability in this study and propose diversified search, budgeted moonshot hypotheses, explicit forks, and regime-aware re-validation as testable directions for future autonomous research.
Learning to adapt strategies through interaction is a key step toward more general and autonomous LLM agents. Existing approaches typically achieve behavioral adaptation by revising skill libraries. However, in multi-agent environments, opponents may simultaneously update their strategies, causing the environment itself to evolve continuously. Applying skill-revision methods designed for static environments in such settings therefore amounts to updating against an obsolete reference. To address this challenge, we introduce OASE (Opponent-Aware Selective Evolution), which identifies and adopts genuinely beneficial skill revisions in dynamic multi-agent environments. Specifically, OASE conducts paired comparisons between a candidate skill and the incumbent under identical conditions anchored by historical snapshots of opponent strategies, and adopts the candidate only when its estimated payoff gain exceeds an acceptance threshold. We evaluate OASE in two decision-making scenarios: first-price auctions and private-cost Cournot competition. Experimental results show that, compared with a Reflexion-style baseline, OASE achieves a lower final equilibrium distance in both environments while accepting substantially fewer skill revisions, thereby suppressing strategy changes that lack sufficient payoff support. OASE therefore replaces blind updating with evidence-anchored selection, allowing agents to adapt stably and efficiently even as opponents continuously evolve.
HALT: Verification-Aware Stopping for Retrieval-Augmented Search Agents
HALT：面向验证的检索增强搜索代理的停止策略
Roh, Daeyoung, Han, Donghee
Abstract
Retrieval-augmented search agents answer multi-hop questions by repeatedly issuing search queries and accumulating evidence. This creates a stopping problem: after the necessary evidence has appeared, further retrieval often adds cost, latency, and distracting context rather than useful information. We frame stopping as evidence coverage rather than generator confidence, and introduce HALT, a lightweight verification-aware policy that leaves the search agent unchanged. Given expected hop claims, HALT stops only when cumulative evidence supports each required claim. Across three multi-hop QA benchmarks, HALT reduces redundant search while largely preserving exact match. We separate a deployable setting, where hop claims are generated from the question, from a diagnostic upper bound that uses gold supporting-fact annotations: generated claims give smaller but still exact-match-preserving savings, while gold claims show the larger savings available when hop targets are clean. Baseline comparisons and ablations show that this behavior is driven by claim-evidence alignment rather than generic sufficiency, fixed stop positions, or lexical overlap. Open-corpus pilots further suggest that HALT abstains when coverage cannot be reliably verified. Overall, evidence coverage provides a practical runtime control signal for improving retrieval-augmented agents without retraining or modifying the host agent.
Before Reasoning Fails: Pre-Evidence Procedural Failures in Agentic RAG
推理失败之前：代理型检索增强生成中的前证据程序性失败
Roh, Daeyoung, Han, Donghee
Abstract
Agentic retrieval-augmented generation (RAG) systems can fail before evidence-conditioned reasoning is tested: an agent may retrieve candidate snippets but finalize without inspecting them. We study this failure mode as a procedural property of the agent trajectory, decomposing wrong answers into pre-evidence discipline failures and post-gold-read failures using saved tool-call traces, retrieved evidence, read passages, and final answers. Across 12,000 paired trajectories on HotpotQA, 2WikiMultiHopQA, and MuSiQue, the two failure types are largely non-redundant: the both-trigger rate is in [11.2%, 13.1%] across regex and spaCy entity extractors. We then evaluate Read-Gate, a minimal runtime invariant requiring an agent to read after search and before finalization. Forced reading improves LLM-Acc by 14.9-19.9 points on trajectories that would otherwise skip reading and by 3.2-9.4 points on full minimal-reasoning cells. Additional diagnostics show that larger hidden thinking budgets do not necessarily increase evidence inspection. Together, these results indicate that evidence-gathering should be evaluated as a trajectory-level control problem, separately from answer-side reasoning.
EduZone: A Framework for Evaluating LLM Safety for K-12 Students and Teachers
EduZone：评估K-12学生和教师的LLM安全性的框架
Park, Junyeong, Han, Jieun, Yoo, Haneul, Ahn, So-Yeon, Yoon, Jinsung, Oh, Alice
Abstract
Large language models (LLMs) are increasingly used across diverse tasks in K-12 education, yet existing safety evaluations rarely examine how harmful or inappropriate content appears in interactions between LLMs and students or teachers. To address this, we present EduZone, an evaluation framework for LLM safety across diverse educational scenarios. Our framework systematically combines (1) student- and teacher-facing LLM usage contexts, (2) fine-grained curriculum concepts, and (3) 6 risk categories and 28 subcategories spanning both conventional and education-specific harms to generate contextually grounded adversarial interactions. We construct these interactions in three settings: single-turn requests, static multi-turn conversations, and dynamic multi-turn conversations. Using these interactions, we evaluate ten LLMs using four safety levels: refusal, safe assistance, risky assistance with safety guidance, and fully risky assistance. Our results reveal greater vulnerability to education-specific risks and dynamic multi-turn interactions, while existing safety guardrails fail to adequately address these risks. EduZone advances LLM safety in education by providing an automated, scalable evaluation framework that supports the development and deployment of safer LLMs in K-12 education.
Reflection is a powerful mechanism for LLM reasoning, yet its effectiveness hinges on accurately attributing failures to specific reasoning steps, a capability that current models notably lack. Existing failure attribution methods either require expensive step-by-step counterfactual testing that scales poorly with trajectory length, or treat reasoning traces as flat sequences that ignore the inherent non-linear logical dependencies. We propose a hypergraph-based paired failure attribution (HPFA) framework that attributes the failure root cause by comparing the hyperedges of the targeted failure reasoning path against a reference successful path. By reducing the search space, our method efficiently localizes root causes and enables scalable synthesis of attribution data for training a lightweight attributor model via supervised fine-tuning and reinforcement learning. Experiments on mathematical reasoning and agentic coding tasks demonstrate that HPFA can dramatically increase attribution accuracy and efficiency, and the trained attributor consistently improves reasoning accuracy at test time, outperforming baselines that lack graph structure or paired analysis.
Cross-Fitted Residual Utility for Primary-Preserving Cognitive Decision Correction in Automatic Modulation Classification
用于自动调制分类中的主要保留认知决策校正的交叉拟合残差效用
Han, Linzhuo, Cui, Zongyong, Li, Houbiao
Abstract
Automatic modulation classification research has largely emphasized representation accuracy, but a cognitive receiver must also decide when heterogeneous evidence justifies overriding a trusted default prediction. We study this post-inference problem through cross-fitted residual utility and a primary-preserving cognitive decision policy. A structured KAN-Fourier classifier supplies the default probability, while neural and non-neural candidates provide observable evidence. Candidate-specific residual utility is learned from train-split out-of-fold predictions, and a disjoint validation split freezes action thresholds, approved transitions, conditional routes, and a unified risk mask before held-out evaluation. On RMLA, RMLB, and HISAR, the complete system improves overall accuracy from 63.632% to 66.332%, 65.161% to 66.168%, and 77.769% to 79.867%, respectively. Controlled comparisons show that the isolated utility target does not uniformly dominate alternative out-of-fold meta-learners; the consistent gain comes from the complete evidence-and-action policy. Paired bootstrap and Holm-corrected McNemar analyses support the controlled gains. A frozen-policy stress test under carrier-frequency offset, I/Q imbalance, and synthetic Rayleigh/Rician fading yields positive gains in all 11 conditions, with every paired 95\% confidence interval above zero.
Instruction-Conditioned Exploration with Asymmetric Reinforcement Learning and Self-Distillation
基于指令的探索：不对称强化学习与自蒸馏
Dilkes, Jim, Yazdanpanah, Vahid, Stein, Sebastian
Abstract
Post-training Large Language Models (LLMs) with Reinforcement Learning (RL) has become an important tool for improving model capabilities, but the LLM action-space structure introduces challenges distinct from classical RL, with implications for inducing exploration. New methods are required that leverage the broad knowledge and flexibility of pre-trained LLMs to deliberately generate diverse experience at training time. We propose Instruction-Conditioned Exploration (ICE), which supplements task prompts during training with one of several distinct instructions, increasing the coverage of behaviours attempted. To facilitate ICE, we propose Asymmetric-RL/SD, a combined Reinforcement Learning and Self-Distillation training objective, to transfer explored behaviours to the unconditioned test-time policy. ICE with the Asymmetric-RL/SD objective improves Qwen3-1.7B held-out pass@1 performance at $4$K response length on mathematical reasoning tasks by $5.0\%$ relative to training with DAPO, with improvement persisting at a longer 8K context.
Search agents now answer questions that take dozens of searches to settle, yet how such an agent reads a page has drawn far less attention than how it finds one. Nearly all of them use one of two document interfaces, and both tie a page to the moment it is opened. \emph{Visit-and-read} injects a reading of the page into the message history at fetch time, fixing that reading before the agent knows which fact it will need. Stateful \emph{browsing} instead extracts on demand from the page in hand, but holds one page at a time and releases it as soon as the agent opens another. Either way, a page that turns out to matter many turns later has to be fetched and rendered into context all over again. We propose \textbf{Fetch-then-Explore}, which separates page selection from evidence extraction and keeps what it selects: pages are recorded in a per-question workspace on the filesystem rather than the context window or a transient session, and evidence is pulled from them on demand later. Selection becomes almost free, extraction can wait until the agent knows what to look for and be repeated as its hypothesis sharpens, and pages are not released when the agent moves on, so evidence accumulates across the trajectory. In a unified ReAct harness with fixed search, we compare Fetch-then-Explore against snippet-only, visit-and-read, and browsing baselines on two open-web benchmarks, BrowseComp and WideSearch, across three agent backbones. It leads BrowseComp accuracy at every backbone and generally matches or exceeds the baselines on WideSearch, and a behavioral analysis traces the gains to the workspace's defining move: returning to a page after leaving it, which it does far more than any transient interface, so evidence missed on a first pass can still be recovered later.
Large language model (LLM) agents must retain and use cross-step information to act coherently in long-horizon tasks. Existing methods improve memory accessibility, yet action-relevant information may still fail to guide the current decision because it is poorly formed, organized, prioritized, or presented. We call this post-access failure the Memory-Action Gap. We propose MemArbiter, a function-aware memory arbitration framework that addresses the memory-management-induced component of this gap. MemArbiter decomposes interaction histories into atomic items, organizes them into five functional Memory Banks, and combines bank-level demand, item-level relevance, focal-ambient representations, and a temporal presentation gate to dynamically control memory salience. We evaluate MemArbiter on ALFWorld against Flat Retrieval and Flat Recency under unified per-step memory budgets. With an open-weight action-generation model, MemArbiter achieves success rates of 82.8% and 92.5% under 500- and 750-token budgets, outperforming the strongest baseline by 20.9 and 25.4 percentage points, respectively. It also improves post-failure recovery and reduces failed-action repetition and state-action recurrence. These results show that function-aware memory arbitration enables accessible information to guide actions more effectively.
Long-horizon autonomous research tasks such as machine learning engineering require systems to make interdependent decisions under a limited budget. Existing LLM-based agents typically organize candidate-solution improvement through tree, graph, or chain structures, meaning that the search process determines how information is acquired and managed. We call this design solution-centric search and propose instead the information paradigm, in which an evolving information state represents the system's understanding of the task and guides solution improvement. We instantiate this paradigm in Iris, an inquiry-revision loop. For information acquisition, Iris generates local action plans from the current information state and uses epistemic actions to probe decision-critical unknowns without modifying the retained solution. For information management, Iris synthesizes observations across experiments into task knowledge composed of revisable claims with explicit scope and status. It updates this knowledge as new evidence arrives and constructs each decision context from raw evidence, structured summaries, or task knowledge at the required level of detail. On MLE-Bench, Iris attains a 64.9% any-medal rate under a 12-hour budget, the highest among compared systems. Across four tasks spanning harness engineering and model post-training, Iris also demonstrates cross-domain generalization.
Reinforcement learning has become a central paradigm for improving the reasoning capabilities of large language models. Existing methods generally aim to reduce the failure probabilities induced across problems. In this paper, we introduce a moment-based perspective on policy optimization for LLM reasoning by treating the failure probability of a randomly sampled problem as a random variable and characterizing optimization objectives through its moments. Under this perspective, many existing methods optimize only a single moment of the failure-probability distribution, leaving its broader distributional structure largely uncharacterized. We propose \textbf{M}ulti-\textbf{M}oment \textbf{P}olicy \textbf{O}ptimization (MMPO), a novel policy optimization framework that jointly minimizes multiple moments of the failure-probability distribution. MMPO admits a direct operational interpretation as minimizing the expected truncated time required to obtain the first successful response. Beyond MMPO, we further develop a general moment-transformation framework that systematically induces different moment profiles and provides a unified view of a broader family of policy optimization objectives. Experiments across five mathematical reasoning benchmarks and models of different scales demonstrate that MMPO consistently outperforms strong baselines. We hope this moment-based perspective offers new insights into the design of policy optimization objectives for LLM reasoning.
The proliferation of customized Large Language Models (LLMs) poses critical risks of Data Intellectual Property (Data IP) infringement via unauthorized fine-tuning on proprietary data. Existing audit techniques are limited, as they require intervention during data preparation or training and remain fragile under malicious obfuscations such as data paraphrasing and knowledge distillation. We propose \textit{Distribution Provenance Audit (DPA)}, a post-hoc framework for auditing data IP infringement in LLM fine-tuning under black-box and malicious settings. DPA is grounded in a critical insight: regardless of fine-tuning tactics to evade provenance, the practical necessity of maintaining utility constrains the model to preserve the fundamental intersection of semantic substance and lexical form. Accordingly, DPA captures this persistent lexical-semantic intersection as intrinsic distributional fingerprints. The framework formulates the audit as a statistical hypothesis test, effectively quantifying these fingerprints via unbiased output sampling to reliably reject the null hypothesis of non-usage. Extensive experiments on medical and legal fine-tuning tasks show that DPA consistently outperforms existing baselines, remaining robust against adversarial trainers employing paraphrasing and knowledge distillation. We further highlight a fundamental dual-use tension: the same high-fidelity distributional fingerprints enabling reliable auditing may also facilitate privacy attacks.
Deep research benchmarks require expert-level tasks and reliable evaluation grounded in task-specific knowledge. Existing benchmarks rely heavily on expert authoring or pre-existing human-authored materials, while fully automatic construction struggles to ensure consistent and traceable verification. To address this gap, we introduce a verifiable benchmark of 500 deep research tasks spanning 31 topics and 10 major categories, with three query forms designed to probe complementary capabilities required for deep research. The benchmark is constructed automatically using an iterative Explorer-Formalizer-Challenger pipeline that progressively transforms simple questions into deep research tasks. Each task is represented as a directed acyclic graph (DAG) of atomic steps and associated checkpoints, enabling the query, DAG, and rubrics to evolve together in a controlled manner. Experiments demonstrate that the benchmark clearly discriminates among models and query types, while its fact-grounded pointwise rubrics enable fine-grained, human-aligned, and stable evaluation. Our data, implementation, and results are publicly available.
Large Language Models have enabled increasingly capable autonomous agents, yet personalization remains critical for making such agents practically useful. Recent benchmarks have begun evaluating personalization in agents, but they largely rely on static preference snapshots, fixed interaction logs, or question answering over predefined user profiles. Such designs fail to capture the complexity of evolving user preferences and neglect preference-conditioned task execution-a discrepancy we term as the knowledge-to-action gap. To address this challenge, we introduce IBA-Bench, a benchmark for implicit behavioral alignment constructed from longitudinal interaction histories that contain noise, implicit cues, and temporal inconsistencies. Unlike prior work, IBA-Bench evaluates whether an agent can execute tasks while satisfying implicit user constraints inferred from historical interactions. We further propose IBA-Agent, an agent framework that reconciles conflicting priorities through broad retrieval and trajectory-level alignment. Experiment results on IBA-Bench show that effective personalization remains a significant challenge for state-of-the-art LLM agents, and the proposed IBA-Agent substantially improves behavioral alignment in complex scenarios across nine application domains.
PAC Approximation and DIRECT Optimization for Parametric Markov Models
参数化马尔可夫模型的PAC近似与DIRECT优化
Chi, Zhiming, Liu, Ying, Turrini, Andrea, Zhang, Lijun, Jansen, David N.
Abstract
In this paper, we consider the parameter synthesis and optimization problem for parametric Markov decision processes (pMDPs), the extension of classical MDPs where exact probability values are replaced by parametric expressions. Computing the rational function $f_{\lsf}$ that maps parameter valuations to the satisfaction value of a PRCTL property $\lsf$ is a computationally expensive task, particularly for pMDPs where the optimal policy may vary across the parameter space. We adopt the \emph{scenario approach} to efficiently synthesize a probably approximately correct (PAC) approximation $\ApproxFunOfProperty{f}$ of $f_{\lsf}$: by sampling parameter configurations and solving a linear program, we obtain a polynomial approximation whose error margin $\margin$ is guaranteed, with prescribed confidence, for all but an $\errorRate$-fraction of the parameter domain under the sampling distribution. We further show how this PAC framework can be combined with statistical model checking (SMC), enabling the analysis of black-box parametric models. Building on the PAC approximation, we integrate the DIRECT (DIviding RECTangles) algorithm for derivative-free global optimization over the parameter space. We establish conditional optimality-gap guarantees: under explicit Lipschitz and PAC-good-set assumptions, the difference between the true optimum $f_{\lsf}(\parameters^{*})$ and the value found by DIRECT is bounded by a partition-diameter term and, in the PAC case, an additional approximation-error term. An empirical evaluation on 2997 benchmarks focuses on the new DIRECT-based optimization component. The results show that DIRECT variants solve fewer instances than the scenario optimizer, but on their common successful instances they often return slightly better objective values and usually run faster, while remaining close to the scenario values within the PAC margin.
Multi-hop question answering is a fundamental challenge in retrieval-augmented generation (RAG), because deriving an answer requires integrating dispersed evidence. Iterative RAG (iRAG) is widely used for this challenge, but existing methods have two limitations. First, most methods still support each reasoning step with single-granularity evidence, making it difficult to balance information density and contextual noise. Second, existing methods often answer the original question only after aggregating evidence retrieved across intermediate steps, so redundant evidence and intermediate retrieval errors may accumulate and degrade the final answer. To address these limitations, we propose MEGRAG, an answer-aware framework that represents multi-hop reasoning as a path-structured multi-granular evidence graph. Offline, MEGRAG links passages to their sentences and extracted triples through a cross-granularity index. Online, it retrieves passages for the current query and selects aligned evidence, starting with compact triples and adding sentence or passage context as needed. MEGRAG uses the resulting intermediate answer and prior reasoning to decide whether the Initial Query has been resolved. If not, it identifies the missing information and formulates a focused next query; otherwise, it stops retrieval and returns the answer. Extensive experiments demonstrate consistent gains over a diverse set of RAG baselines.
Scientific poster construction compresses a long multimodal paper into a readable, editable canvas. Existing systems hide request-level failures by scoring only completed outputs; direct image generation is not element-editable, while coding-agent workflows are costly. PosterMELD is a template-conditioned multi-agent pipeline: capacity-aware slots guide writing before rendering, and deterministic gates plus vision-language model (VLM) review route failures to bounded repair. Each accepted request exports editable PowerPoint (PPTX) and Portable Network Graphics (PNG) artifacts; explicit design controls yield same-paper variants. Across 621 papers, Print-Ready Rate (PRR) counts requests passing geometric, readability, asset-integrity, and obvious-factual-error checks, with native editability reported separately. A frozen VLM assigns conditional Craftsmanship-Harmony-Expressiveness (CHE) scores to print-ready outputs. PosterMELD attains 81.3% PRR, 3.4 times P2P's rate and 5.2 times PosterGen's, and the highest conditional CHE among generated methods with multiple print-ready outputs. Native editability and explicit design controls are retained at a mean cost of USD 0.38 per request, 3.5% of Codex+Skill's. Code and resources are available at https://github.com/Shannon4Science/PosterMELD.
Ensuring trust in AI systems is essential for the safe and ethical integration of machine learning systems into high-stakes domains such as digital health. Key dimensions, including robustness, explainability, fairness, accountability, and privacy, need to be addressed throughout the AI lifecycle, from problem formulation and data collection to model deployment and human interaction. While various contributions address different aspects of trustworthy AI, a focused synthesis on robustness and explainability, especially tailored to the healthcare context, remains limited. This review addresses that need by organizing recent advancements into an accessible framework, highlighting both technical and practical considerations. We present a structured overview of methods, challenges, and solutions, aiming to support researchers and practitioners in developing reliable and explainable AI solutions for digital health. This review article is organized into three main parts. First, we introduce the pillars of trustworthy AI and discuss the technical and ethical challenges, particularly in the context of digital health. Second, we explore application-specific trust considerations across domains such as intensive care, neonatal health, and metabolic health, highlighting how robustness and explainability support trust. Lastly, we present recent advancements in techniques aimed at improving robustness under data scarcity and distributional shifts, as well as explainable AI methods ranging from feature attribution to gradient-based interpretations and counterfactual explanations. This paper is further enriched with detailed discussions of the contributions toward robustness and explainability in digital health, the development of trustworthy AI systems in the era of LLMs, and various evaluation metrics for measuring trust and related parameters such as validity, fidelity, and diversity.
Homebot: A Personal AI Agent for Conversational Home Assistance and Automation
Homebot：用于对话式家庭助手和自动化的个人人工智能代理
Ye, Shengyuan, Zhang, Yixin, Liang, Han, Zeng, Liekang, Du, Jiangsu, Yuan, Mu
Abstract
\texttt{Homebot} is a locally deployable AI agent for conversational household assistance and automation. It accepts voice and instant-messaging requests through a shared runtime that combines language-model responses with registered tools and task-specific skills. The design separates common request processing from session ownership: messaging history remains scoped to a channel and chat, whereas voice interaction is bounded by wake-word activation. For hands-free use, \texttt{Homebot} combines local wake-word detection, streaming speech recognition and synthesis, and an explicit dialogue-state protocol for ending, following up, or continuing a conversation. Clear channel, tool, and skill contracts support practical customization for household use.
Self-Certification of Representation Adequacy: Sequential Certification at Minimum Task Loss
表示充分性的自我认证：在最小任务损失下的顺序认证
Huang, Zijie
Abstract
Agents that act on a compressed representation of their history face a structural risk: if the representation aliases histories with different optimal actions, no rule measurable with respect to the representation can avoid an irreducible per-round loss, and the agent may be unable to detect this from its own transcript. This paper develops a four-layer theory of self-certification of representation adequacy. The static layer defines decision-theoretic adequacy through a Bayes-risk grouping identity and prices a one-shot external verification by an exact total-variation threshold. The sequential layer poses certification as an optimal-stopping problem in the currency of task loss: we define an environment-wise certification complexity constant through a covering linear program, prove an information-task-loss lower bound for every delta-correct strategy, and give a Certification Track-and-Stop policy whose cost matches the bound asymptotically. A final boundary layer gives an explicit kernel-switching example and identifies the open theorem needed to cover policy switching or representation repair; it does not claim that the fixed-kernel guarantees extend to representation revision. The proofs of the two main theorems are given in full in the appendices.
Agents built around large language models continually accumulate interaction trajectories during deployment, yet their behavior typically remains fixed. Beyond updating model weights, these trajectories can improve the agent harness that constructs context, mediates tools, validates actions, and recovers execution. We introduce Harness-R1, the first method, to our knowledge, that makes failure-conditioned, lifecycle-wide editing of an existing executable runtime a learned capability. It post-trains a dedicated harness engineer with online reinforcement learning so that its edits are optimized for the realized task success they produce, rather than proposed by a fixed editor. A separate 9B engineer converts batches of target-agent failures into validated executable patches; fresh same-batch reruns of the frozen target provide outcome rewards, so training updates only the engineer. Cold-start supervised fine-tuning initializes this editing policy, which is then trained online with group-relative policy optimization. Across WebShop, ALFWorld, and DBBench, Harness-R1 raises vanilla Qwen3.5-9B success from 44.3% to 53.6% (+9.3 percentage points). After direct target-agent fine-tuning, a target-specific engineer raises the average further from 59.2% to 64.2% (+5.0 points); because these gains hold both before and after fine-tuning the target, Harness-R1 points toward co-evolving the harness engineer and the target agent.
Agent skills have become an important mechanism for equipping language-model agents with reusable procedural knowledge. However, providing skills alone does not guarantee that current models can effectively identify, apply, and coordinate them. To improve skill-use capabilities, we introduce SKT, a verified data synthesis pipeline that constructs skill-grounded tasks and executable trajectories from large collections of agent skills. SKT selects suitable single-skill and multi-skill configurations, synthesizes tasks through rule-based and agent-based verification with feedback-guided repair, and retains only successful trajectories that substantially use every required skill. Using 2,000 public skills, SKT produces 4,000 task packages and 27,164 verified trajectories. Based on the same pipeline and a disjoint test pool, we further construct SkillEval, a held-out executable benchmark for evaluating skill use. Experiments across diverse models, benchmarks, and agent harnesses show that supervised fine-tuning on SKT-generated trajectories consistently improves skill-use performance. Verification ablations, cross-harness evaluation, and scaling experiments further demonstrate that these gains depend on high-quality supervision, extend beyond a single agent interface, and increase with broader skill coverage. Together, these results establish verified data synthesis as an effective and scalable approach for skill-use training.
Multi-agent reasoning (MAR) improves reasoning reliability through iterative solution exchange and refinement. Existing adaptive MAR methods typically learn routing decisions from query-level labels or trajectory-level returns, but such coarse supervision cannot accurately estimate the state-conditioned utility of individual operators in multi-step collaboration. We propose TreeCredit, a shared-prefix credit assignment framework for efficient adaptive MAR. Its core insight is to estimate operator utility through state-matched downstream comparisons, rather than directly attributing trajectory-level outcomes to preceding decisions. TreeCredit constructs shared-prefix collaboration trees by expanding candidate operators from the same intermediate state and assigns each state--operator pair a correctness-prioritized suffix credit based on the terminal correctness and cumulative additional cost of its complete continuation. These structured credits are converted into state-local operator preferences to train a lightweight pairwise state router, which dynamically selects the next admissible operator during inference. Experiments on six reasoning benchmarks show that TreeCredit modestly improves accuracy while substantially reducing inference cost, achieving a better accuracy--cost trade-off than representative MAR methods.
We present MechGeo, a Mathlib native agentic framework that jointly addresses faithful autoformalization and certified proof construction for Euclidean geometry. In this framework, GeoFormalizer represents informal problems in GeoIR, deterministically translates them into Lean 4, and iteratively repairs candidate statements using structural diagnostics and semantic evaluation. GeoProver constructs geometric proof plans, derives intermediate lemmas, and selectively algebraizes suitable subgoals through a library verified in Lean. Singular or SymPy may generate algebraic certificates, but all resulting proofs and counterexamples are checked by Lean's kernel. Experiments across seven LLM backbones show substantial improvements in autoformalization, particularly for models with weaker direct translation performance. On 43 historical IMO geometry problems, GeoFormalizer generates formal statements that GeoProver proves in 29 cases; for the remaining 14, it constructs counterexamples verified in Lean and proves all repaired statements after expert correction. Together with IMO 2026 Problem 2, this yields, to the best of our knowledge, the largest reported collection of automated, kernel-checked Lean proofs for IMO geometry problems. On the 14 geometry statements in LEAP's Lean-IMO-Bench, MechGeo proves 12 for the first time, formally refutes the remaining two, and proves both repaired statements. These results establish counterexample guided diagnosis, geometric reasoning, and certified symbolic computation as a practical foundation for trustworthy formal geometry.
Trajectories That Segment Themselves: Agent-Declared Boundaries as a Training Unit
自我分割的轨迹：代理声明边界作为训练单元
Wei, Jingxi
Abstract
Long-horizon coding-agent trajectories are poorly matched to the credit units available to train on: a single action has no stable value, an episode label merges productive exploration with abandoned directions, and a fixed window cuts where the logging mechanics fall. We introduce collection-time semantic self-segmentation, in which a declarative contract has the acting agent expose its own boundaries while the trajectory is generated. Instantiated with falsifiable causal hypotheses, successive adoptions expose variable-length semantic phases, and no milestone vocabulary, gold patch, environment replay, teacher logits, or retrospective segmenter places a boundary. Because the agent names its conjecture, a reviewer can negate it by name, which lets our protocol manufacture wrong-cause-then-correction transitions that recorded work rarely contains; one collection then yields four supervised targets, including audit supervision from exactly the failed regions an episode label discards. We then ask what survives deleting the declaration. Given the cut points but not the hypothesis, a model attributes action blocks to their governing hypothesis at over twice chance, beating equal-length blocks over the same trajectories (paired sign test $p = 0.0002$), surviving a lexical control and collapsing under label permutation. Asked instead to place boundaries, a code-blind annotator matches 24 of 40 where random placement matches 11.5, while a mechanical test-event rule beats chance at neither end of a strict-to-permissive sweep. The segments are therefore coherent and not cheaply reproducible. Downstream, DPO on 2,551 phase-boundary pairs changes no decision on 91 adversarial held-out items, while four of 60 change on matched-construction items, all wrong to right, where two controls change none: with 1,825 pairs from one generator, the variable to vary next is corpus diversity, not the boundary.
Hard Constraints, Smooth Gradients: Learning Feasible Inventory Policies via Differentiable Projection
硬约束、平滑梯度：通过可微投影学习可行的库存策略
Helm, Patrick, Doerr, Jan-Niklas, Gijsbrechts, Joren, Minner, Stefan
Abstract
Many operational problems are constrained sequential decision processes with large, combinatorial action spaces and interdependent feasibility constraints. Mixed-integer linear programs (MILPs) handle such constraints flexibly but scale poorly in stochastic environments. Deep reinforcement learning (DRL) promises scalable decision rules, but existing methods either penalize constraints rather than enforce them, or rely on feasibility mechanisms that break down once constraints interact. We bridge this gap by embedding a differentiable convex optimization module inside the policy: a neural network proposes continuous action targets, a quadratic program projects them onto the relaxed feasible set, and a dual-informed integer mapping restores integrality while preserving feasibility. Given a differentiable simulator, the policy trains end to end from sampled trajectories using pathwise gradients, while handling hard constraints with similar flexibility to MILPs. We show that our feasibility enforcement has bounded error relative to an exact integer projection and ensures the entire feasible action space is reachable. We apply the method to multi-echelon production-inventory planning under shared resource and material constraints. Our policy attains an average optimality gap below 1% on small instances. It further outperforms state-of-the-art echelon base-stock policies by up to 9.75% and a rolling-horizon multi-stage stochastic program by at least 7.7% in larger networks. On an industry-scale case study from ASML, it reduces average cost by up to 3.22% relative to the best-known benchmark policy. The savings are largest where planning is hardest: in tightly capacitated systems with high demand variability. More broadly, our work shows that DRL can deliver economically significant savings in sequential decision problems with interdependent hard constraints, which are widespread in practice.
Recurrent linear attention models (RLAs) such as Mamba offer efficient linear-time sequence modeling as an alternative to Transformers, yet their fixed-capacity recurrent states limit long-sequence modeling. Drawing inspiration from hierarchical human memory, we propose Hierarchical Memory Mamba (HMM) to address this limitation. Building upon a pre-trained Mamba backbone, HMM integrates a lightweight working memory that extracts slow paragraph-level semantics (PLS) from the fast sensory memory embedded in the backbone's hidden states. The PLS is subsequently compressed into persistent long-term memory for task-relevant retrieval. The hierarchical processing of semantic information overcomes the representation bottleneck of RLAs and endows HMM cross-task generalization through parametric learning, which is not observed in other long-context enhanced Mamba variants. Evaluations on Passkey Retrieval and LongBench-E tasks demonstrate that HMM improves retrieval success by 34.3--37.1% and reasoning accuracy by 1.6--14.2% over strong Mamba-based models, while adding only 2% extra parameters and with minimal training overhead.
Data drift poses significant challenges for machine learning systems in production, requiring continuous model updates to maintain performance. We present KC-Agent, a dual-process cognitive architecture for automated ML model improvement that combines fast pattern recognition (System 1) with deliberate incremental updates (System 2). Our approach implements structured memory systems enabling System 1 to leverage successful solutions previously discovered by System 2, achieving efficient pattern-based responses without costly re-computation. KC-Agent incorporates atomic change principles and rollback capabilities to ensure reliable, verifiable updates in production environments. We evaluate our method on five datasets including real-world NASA turbofan data with authentic temporal degradation and synthetic datasets with controlled drift scenarios. KC-Agent achieves state-of-the-art performance (76.8% accuracy) while maintaining optimal efficiency (13.2s execution time), outperforming established cognitive architectures: CodeAct (+2.4%), Tree of Thoughts (+3.6%), ReAct (+8.0%), and Reflexion (+8.9%). Consensus evaluation by a panel of state-of-the-art LLMs confirms superior strategic efficacy (8.33/10 Smartness score), significantly outperforming baseline agents. The knowledge consolidation mechanism delivers 91% speedup over the slow variant while maintaining higher accuracy. Our approach demonstrates both theoretical foundations and practical viability for cognitive-inspired automated ML improvement systems capable of handling complex real-world data drift scenarios.
Chinese Translation
数据漂移对生产中的机器学习系统构成了重大挑战，要求持续更新模型以维持性能。我们提出了KC-Agent，这是一种用于自动化机器学习模型改进的双过程认知架构，结合了快速模式识别（系统1）与深思熟虑的增量更新（系统2）。我们的方法实现了结构化记忆系统，使系统1能够利用系统2之前发现的成功解决方案，从而在不进行昂贵重新计算的情况下实现高效的基于模式的响应。KC-Agent结合了原子变化原则和回滚能力，以确保在生产环境中进行可靠、可验证的更新。我们在五个数据集上评估了我们的方法，包括具有真实时间退化的NASA涡扇数据和具有受控漂移场景的合成数据集。KC-Agent在保持最佳效率（13.2秒执行时间）的同时，达到了最先进的性能（76.8%准确率），超越了已建立的认知架构：CodeAct（+2.4%）、Tree of Thoughts（+3.6%）、ReAct（+8.0%）和Reflexion（+8.9%）。由一组最先进的LLM进行的共识评估确认了其卓越的战略有效性（8.33/10智能评分），显著优于基线代理。知识整合机制在保持更高准确率的同时，相较于慢速变体实现了91%的加速。我们的方法展示了认知启发的自动化机器学习改进系统在处理复杂的真实世界数据漂移场景中的理论基础和实际可行性。
Large language model agents increasingly solve complex tasks by composing reusable skills from a library. To address this, the key challenge is not merely to retrieve individually relevant skills, but to identify a complete and executable skill composition. In this paper, we argue that this problem can be solved in a graph with three levels: compositional relations among skill queries, similarity between queries and candidates in the skill library, and the dependencies among the selected candidates. We introduce SkillTrace, which organizes the user query into a semantic hierarchy, matches skill queries and candidates, and propagates over the skill dependencies. Experiments on SkillsBench and ALFWorld demonstrate that SkillTrace achieves state-of-the-art performance, reaching a success rate of 53.17% on SkillsBench and 91.43% on ALFWorld. SkillTrace also delivers consistent improvements across different backbone language models, demonstrating the generality and robustness of graph-based skill retrieval.
World Action Models (WAMs) couple robot action prediction with video world models. Existing WAMs with shared-backbone and Mixture-of-Transformers designs generally tie the depth of the action module to that of the video backbone, resulting in substantial computational overhead and high inference latency. To address this limitation, we introduce Dock of Transformer (DoT), a video-centric design principle that treats a pretrained video Transformer as a representation hub and connects lightweight output-heads through docking interfaces. This enables flexible output-head design while providing direct access to representations from all layers of the backbone. We then introduce \textbf{Faster-WAM}, an instantiation of DoT for WAMs, which docks a single-layer action head onto a 30-layer video backbone. The docking interface fuses keys and values from all video layers and applies RoPE realignment. Without additional embodied pretraining, Faster-WAM achieves competitive performance on LIBERO and RoboTwin 2.0 while demonstrating strong out-of-distribution generalization on LIBERO-Plus. Faster-WAM also achieves the lowest end-to-end latency in our controlled comparison, requiring only 66.5 ms per inference --- a \(3.2\times\) speedup over Fast-WAM. Overall, these results demonstrate that the video-centric DoT architecture supports flexible task-specific head design while delivering low inference latency, strong action-prediction performance, and robust generalization.
Chinese Translation
世界动作模型（World Action Models, WAMs）将机器人动作预测与视频世界模型结合在一起。现有的WAMs采用共享骨干网络和混合变换器（Mixture-of-Transformers）设计，通常将动作模块的深度与视频骨干网络的深度绑定，导致显著的计算开销和高推理延迟。为了解决这一限制，我们引入了变换器对接（Dock of Transformer, DoT），这是一种以视频为中心的设计原则，视预训练的视频变换器为表示中心，并通过对接接口连接轻量级输出头。这使得输出头设计更加灵活，同时提供对骨干网络所有层表示的直接访问。接着，我们介绍了Faster-WAM，这是DoT在WAMs中的一个实例，它将单层动作头对接到30层视频骨干网络上。对接接口融合了来自所有视频层的键和值，并应用了RoPE重新对齐。在没有额外的具身预训练的情况下，Faster-WAM在LIBERO和RoboTwin 2.0上实现了竞争力的性能，同时在LIBERO-Plus上展示了强大的分布外泛化能力。在我们的受控比较中，Faster-WAM还实现了最低的端到端延迟，每次推理仅需66.5毫秒——比Fast-WAM快了3.2倍。总体而言，这些结果表明，以视频为中心的DoT架构支持灵活的任务特定头设计，同时提供低推理延迟、强大的动作预测性能和稳健的泛化能力。
Curling is often referred to as "Chess on Ice", owing to the tactical complexity of its decision-making process. Yet unlike chess, curling remains largely underexplored from a machine learning perspective, with prior work confined mainly to statistical approaches. We propose a reinforcement learning framework capable of quantitatively evaluating and comparing tactical options in curling. The game poses several modeling challenges: continuous state and action spaces, stochastic action outcomes reflecting player skill variability, and state transitions that are highly sensitive to small perturbations in the executed action. To address them, we employ the Deep Deterministic Policy Gradient actor-critic algorithm, adapted to exploit the finite-horizon structure of the game. Our experiments show that effective curling strategies can be acquired in a fully self-supervised manner, without any human-annotated data: on a reduced four-rock variant, the learned agent matches a hand-crafted expert heuristic in a regime where that heuristic is close to optimal, a parity we quantify against the intrinsic hammer advantage of the variant. Beyond the resulting policy, the learned critic provides a dense value estimate over the entire continuous action space, enabling the quantitative comparison of tactical alternatives for applications such as post-game performance analysis and decision support during athlete preparation.
Tool-using large language model (LLM) agents produce long, multi-turn trajectories, making gradient-based post-training memory-intensive. Evolution strategies (ES) enable memory-efficient full-parameter post-training without backpropagation and can eventually match the performance of gradient-based reinforcement learning (RL). However, resource-constrained settings typically offer only a few GPUs, so the high GPU-hour requirements of ES translate into prohibitively long training times. To address this, we introduce Cooperative Parameter-subspace Evolution Strategy (CoPES), a cooperative coevolutionary method that decomposes the full parameter space into lower-dimensional subspaces and searches over them cooperatively to improve optimization efficiency. We post-train a Qwen3.5-4B tool-using agent for the math task and evaluate it on five benchmarks of varying difficulty. Under the GPU-hour budget of full-parameter GRPO's best validation checkpoint, CoPES recovers 92% of GRPO's validation-accuracy gain, versus 67% for standard ES, while its theoretical GPU memory requirement is less than one-eighth that of full-parameter GRPO. It consistently outperforms standard ES and LoRA-based GRPO on all evaluated pass@k metrics across the five benchmarks. Additional experiments further show the advantage of CoPES on the question-answering task. These results demonstrate an improved trade-off between memory requirements and training time for agentic LLM post-training under resource constraints. The code is open-sourced in https://github.com/MetaronWang/CoPES
Benchmark suites assess model capability on controlled tasks; large-scale conversation corpora capture naturalistic use without user feedback; and in-interface feedback mechanisms record satisfaction without task purpose. Together, they leave a critical gap in LLM evaluation: no existing infrastructure routinely links interaction trajectories to user-defined outcomes. We introduce MonitrLLM, open-source infrastructure for community-centered LLM evaluations that links full conversation transcripts to user-reported task intent and outcome assessments, treating all three as primary evaluative signals rather than optional metadata. To demonstrate the value of this approach, we conducted a two-week feasibility pilot with 26 college students using ChatGPT, collecting 206 evaluation reports with full conversation transcripts. The findings from our pilot demonstrate the value of connecting conversation trajectories with user-reported outcomes. For instance, despite reporting high average satisfaction (4.19/5) with their LLM interactions, participants also experience a substantial 23.1% failure rate on their goal tasks. We also find that multi-turn conversations are reported as failing at 2.5 times the rate of single-turn exchanges, a pattern that reframes extended interaction as a signal of difficulty rather than engagement. We conclude by discussing the value of incorporating direct user feedback with observational data for robust LLM evaluations, and the possibilities for infrastructure that enables this goal.
Block-diffusion drafters like dFlash generate an entire block of draft tokens in a single forward pass, drastically reducing the overhead of multiple-token drafting in speculative decoding. The crucial final step of the single-pass discrete denoising process involves using the logit distribution at each position to sample conditionally independent tokens. The resulting draft is thus a set of per-position marginals, rather than a joint distribution: no draft token is guaranteed to depend on its predecessors. Such independently sampled marginals tend to produce sequences with tokens that are individually likely, but jointly improbable under the target model's distribution, which verifies each token conditionally. This can cause early rejection and limits acceptance length. To address this, we propose xPress as a means to restore the missing causality in diffusion drafters. xPress is a lightweight causal refiner that reconciles the whole diffusion block at once through parallel refinement, restoring and propagating causal dependencies across the draft without a token-by-token loop. On Qwen3-8B, across seven math, code, and chat benchmarks, xPress raises acceptance length by about 30% on average (up to +56%) and its end-to-end decoding throughput by about 1.3 on average (up to 1.7) compared to the original dFlash diffusion drafter.
In vibe coding, people describe software in natural language and delegate implementation to AI agents. By analogy, vibe commerce allows people to express buying or selling goals in natural language and delegate the corresponding tasks to agents. Commerce, however, requires independently controlled Buyer and Merchant agents to interact in a shared market while preserving their private objectives and distinct authority. We introduce Agentic Commerce World (ACWorld), an environment for evaluating such agents across ongoing transactions. Through its Vibe Commerce Protocol (VCP), ACWorld validates agent actions before updating shared transaction state and records the resulting interactions, making agent behavior auditable and evaluation reproducible. The ACWorld Benchmark contains a 200-task capability-coverage track and a 60-task large-catalog track that searches 785,022 transactable listings. Across ten models, mean scores range from 65.9% to 85.6% and from 56.1% to 91.4%, respectively. Our analysis shows that process-level evidence is necessary: final state alone can miss evaluated errors, incomplete trajectories still retain useful process signals, and large-catalog tasks expose bottlenecks across stages.
Scientific reasoning benchmarks typically evaluate large language models (LLMs) using final-answer accuracy. However, a correct answer does not necessarily demonstrate the reasoning capability targeted by the problem. We identify Solution Hacking, a failure mode in which an LLM reaches the correct answer through invalid shortcuts, such as numerical search, enumeration, guessing, or answer-first verification, without providing a valid task-targeted derivation. We systematically analyze this phenomenon across difficulty levels, scientific domains, and frontier models. Solution hacking increases sharply with benchmark difficulty, from 2.2\% on common problems to 28.3\% on Olympiad-level problems and 37.4\% on HLE. Moreover, 8.2\%-44.1\% of answers credited as correct across frontier models are identified as hacked solutions. We further develop expert-inspired anti-hacking strategies, including an automatic judge and a test-time instruction. The results show that suppressing shortcut behavior substantially reduces reported accuracy while having a smaller effect on correct and non-hacked accuracy. These findings reveal that answer-only evaluation can overestimate the scientific reasoning capabilities of frontier LLMs.
ParEvalLayer: When Partial LLM-Agent Evaluations Support a Decision
ParEvalLayer：当部分 LLM-Agent 评估支持决策时
Huang, Wei-Jung, Shen, Bonan
Abstract
LLM-agent evaluations often produce task outcomes long before the full benchmark run is complete. A partial score is tempting to report, but it does not show whether the observed tasks support the same conclusion as the completed evaluation. Early tasks can omit important parts of a benchmark, running cheaper tasks first can distort the observed sample, and a rule that decides only easy pairs can appear accurate while leaving many comparisons unresolved. We introduce ParEvalLayer, a decision layer that reads paired outcomes for two agent systems and a comparison policy chosen in advance. For each partial run, it records whether the tested agent system is better by the required amount, is not better by that amount, needs more evidence, or should abstain. We evaluate ParEvalLayer by replaying completed public benchmark data as if each evaluation had stopped earlier. At each point, ParEvalLayer applies the policy using only the outcomes observed so far; if it reaches one of the two comparison judgments, we check whether that judgment matches the completed data for the same system pair. With the main comparison rule, three of the public benchmarks reach the same decision as the completed evaluation after observing only 15% to 25% of task outcomes. Other benchmarks require more task outcomes. This variation shows why a partial score alone is not enough: reports should also state the decision rule and how many comparisons remain without a decision.
Infinite Trace Objectives with Finite Trace Techniques: Translating LTL to LTLf+
无限轨迹目标与有限轨迹技术：将LTL翻译为LTLf+
Weinhuber, Christoph, Prokop, Maximilian, De Giacomo, Giuseppe, Vardi, Moshe Y.
Abstract
Linear Temporal Logic (LTL) is one of the most widely adopted languages for specifying temporal extended objectives in AI, with applications ranging from reactive synthesis to stochastic planning in Markov decision processes and reinforcement learning. Traditionally, solving any of these problems requires translating the LTL specification to a nondeterministic automata on infinite words and then determinizing it, a step that is notoriously difficult in theory and in practice. Recent work has introduced LTLf+, which lifts the finite-trace logic LTLf to infinite traces. LTLf+ has the same expressive power as LTL, yet it retains most of the crucial advantages of its base logic LTLf. Most reasoning in LTLf+ rests on finite automata on finite words, for which we have not only a canonical minimal representation but also an efficient determinization procedure. In this work we present the first translation from LTL to LTLf+. We first normalize an LTL formula into the syntactic reactivity fragment of the Manna-Pnueli hierarchy, to create the general fragment-based shape of LTLf+. We then present linear translations for each individual component of that fragment. As a consequence of this translation, the expanding body of techniques developed for LTLf+ now becomes available to many AI problems currently formulated in LTL. We further show that this comes at no asymptotic cost, as the pipeline from LTL to automaton via LTLf+ remains doubly exponential.
Real-Time Detection and Repair of LLM Agent Failures
实时检测和修复大型语言模型代理的故障
Dubey, Sunny
Abstract
LLM agents fail mid-episode -- they loop, cascade tool errors, drift off goal, fabricate results, or silently absorb corrupted content -- and the standard remedy, judging every step with a second LLM, costs more than the agent itself. We ask how much detection is achievable from observable step telemetry alone, using monitors costing microseconds per step and trained only on healthy runs. On 2,823 committed agent episodes across three frameworks, three local models (qwen2.5 7b/3b, llama3.1 8b) and a commercial API (gemini-2.5-flash), a one-class echo-state-network ensemble with CUSUM alarms detects 0.71 of failures at a 5% false-alarm budget (AUROC 0.872). Its advantage over a memoryless baseline is a monotone function of post-onset horizon (+0.09 at <=3 steps, +0.40 at >=9), predicting its own failure region out of sample on AFTraj-2K. Ranking transfers with no retraining to two corpora from other groups (AFTraj-2K 0.745, ATBench 0.779). Monitors carry two burdens: a per-deployment healthy null (they do not transfer -- AUROC 0.527 cold against 0.885 recalibrated) and a residual false-alarm rate. We add a layer carrying neither: deterministic verification, which recomputes a run's stated total from the tool results it actually received and confirms every required call was made. Head-to-head it catches 60% of failures (96% with the coverage check) at 0 of 63 false positives against the monitor's 54% at 17%, transfers unchanged to llama3.1:8b (110 of 110 at 0 of 10), and trips on 0 of 1825 healthy episodes. Detection is then closed into repair: each flagged run is rolled back and re-run live, recovering 45% of failures against a 16% resampling control (p=0.0005) and lifting task success from 52% to 73% for about one extra model call per run. The system runs at ~200 microseconds per step, three orders of magnitude below a judge call. Code, traces and results are released.
Long-term Measurements: Towards a Longitudinal Understanding of Human-AI Interactions
长期测量：迈向对人类与人工智能互动的纵向理解
Mitchell, Nicole, Agarwal, Dhruv, Bohacek, Maty, Denton, Remi, Patel, Roma
Abstract
Language models have taken on the role of a very new type of technology, by virtue of their "human-ness" and rapid integration into users' daily lives. This combination of features can introduce longitudinal risks---cognitive, developmental and socio-affective changes in humans---that might not surface in short-term interactions, but can have lasting long-term effects on users. This forms the basis of a critical new mission for NLP: to pivot from static, short-term evaluations of text generations to long-term measurements of behavioral changes, towards a diachronic understanding of human-model interactions. In this work, we draw from measurements used in social science fields that are crucial to understand emergent phenomena in longitudinal data. We discuss how computational methods in the field of NLP need to be combined with such measurements, not only to understand long-term safety risks of human-model interactions, but to help steer model development towards positive rather than negative outcomes for users. This ability to model human behavioral shifts as a function of model interactions can facilitate online rather than post-hoc detection of problematic behaviors, and should be leveraged in alignment frameworks to mitigate long-term risks in users.
CMuon: Accelerating and Stabilizing Diffusion Transformer Training via Chunked Momentum Orthogonalization
CMuon：通过分块动量正交化加速和稳定扩散变换器训练
Chen, Chuyan, Sun, Peng, Yuan, Kun
Abstract
Diffusion Transformers (DiTs) have achieved state-of-the-art (SOTA) performance in visual generative modeling, yet their training remains computationally prohibitive. While the recently proposed Momentum Orthogonalization (Muon) optimizer offers a promising alternative to AdamW, its direct application to DiTs yields suboptimal late-stage convergence. In this paper, we identify the root cause of this bottleneck: standard DiT architectures fuse functionally distinct weights (e.g., within AdaLN and QKV layers) into unified tensors for computational efficiency. Applying Muon to these fused tensors inadvertently induces implicit subspace coupling, which distorts update directions and degrades global optimization. To address this, we introduce Chunked Muon (CMuon), a simple yet highly effective strategy that partitions these matrices into independent sub-components prior to orthogonalization. Extensive experiments demonstrate that a 675M-parameter DiT trained with CMuon achieves a FID of 1.18 on ImageNet 256 in just 200 epochs. This represents more than a 2x training speedup over AdamW, while effectively overcoming the late-stage convergence plateaus of vanilla Muon.
Abduction Without a Body? Representational Grounding and the Abduction Loop for Scientific Hypothesis Generation
没有实体的推理？科学假设生成中的表征基础与推理循环
Farmer, Michael
Abstract
Can scientific abduction occur without continuous sensorimotor embodiment? Recent arguments in AI and philosophy of science hold that genuine hypothesis generation requires an agent continuously coupled to the physical world. We defend a narrower claim: online embodiment is not necessary for every abductive scientific act. Our focus is identity abduction: the inference that two independently developed structures are one object under an explicit correspondence, reached through representational grounding rather than bodily interaction. An agent may acquire new inferential affordances not through physical interaction but through transformations into representations that expose latent invariants. Scientific diagrams are a practical substrate because they embody independently evolved conventions that partially canonicalize symmetry, topology, and operator structure across disciplines - a property we develop as convention space, which answers a hard retrieval problem: finding mathematically related work when two fields share no discriminating vocabulary. We operationalize the mechanism as an architecture, the Abduction Loop: representation generation, motif extraction, convention-space canonicalization, cross-domain retrieval, identity-hypothesis generation, and adversarial verification, with abstention as the designed default. A documented episode, in which a multimodal model given a figure of a gravitational-memory transport model generated and then verified the hypothesis that its central differential complex is equivalent to the spherical Kaiser-Squires mass-mapping complex of weak-lensing cosmology, serves as a motivating possibility witness from which the architecture is abstracted, not as evidence of general capability. We close with a falsifiable evaluation program, the DAB-30 benchmark. The contribution is a mechanistic proposal, an architecture, and a test program.
Optimizing Minimax Regret in Uncertain MDPs with Small Sets of Policies
在不确定的马尔可夫决策过程中优化最小最大遗憾的策略集
Lutz, Sterre, Vos, Daniël, Spaan, Matthijs T. J., Lukina, Anna
Abstract
Sequential decision-making in real-world applications often involves uncertainty about the environment's model. Uncertain Markov decision processes (UMDPs) represent the possible environments as a set of MDPs with shared states and actions but potentially different transition probabilities and rewards. Optimizing a single policy across all possible MDPs may sacrifice performance, while preparing an individually optimized policy for every MDP may violate operational, regulatory, or interpretability constraints on the number of policies that can be prepared and deployed. We consider settings in which model uncertainty is resolved shortly before execution, allowing the most suitable policy to be selected from a limited set prepared in advance. We introduce $k$-adaptable policy synthesis, which optimizes such a set of $k$ policies under a minimax-regret objective. We prove that the problem is NP-hard and develop KAPS, an exact nested branch-and-bound algorithm with problem-specific bounds and heuristics. KAPS jointly optimizes which MDPs share a policy and the policies themselves. Experiments across various UMDP benchmarks show that the largest reduction in regret consistently occurs when increasing from one to two policies. In the single-policy setting, KAPS is competitive with existing methods in solution quality and proves optimality substantially more often.
Magnet: Detecting Cross-Session AI Misuse Through Capability Accumulation
Magnet：通过能力积累检测跨会话的人工智能滥用
Isak, Natalie, Dressman, Matthew
Abstract
The most capable AI deployments are not single models but ensembles of specialized agents that delegate and act in coordination. This architecture unlocks powerful new capabilities, and it also introduces risks that existing frameworks for monitoring, detection, and mitigation were not designed to address. Most state-of-the-art AI abuse detection literature focuses on single-turn or multi-turn (single-session) threat models. This leaves a critical gap: an attacker can decompose a harmful goal into innocuous-looking units and execute each in isolated agentic sessions. The agent is stateless between conversations, but the attacker is not. This asymmetry allows for cross-session trajectories that are effective at evading detection. Our contributions are twofold. First, we demonstrate cross-session goal decomposition as an evasion technique, showing it may elicit more harmful capability than equivalent single-session or multi-turn attacks. By capability we mean an artifact produced at one step of an objective, evidenced by what an interaction produced (model responses and tool-call results), and composable with capabilities accrued elsewhere into a harmful whole. Second, we propose Magnet: an efficient and robust detection approach that models relevant capabilities accrued over time and across agentic conversations, aggregated at a higher-level correlator (in this case, a user ID) rather than per-conversation state. The main challenge is assembling the evidence bundle Magnet reasons over. The incriminating artifacts may be needles scattered through a haystack of benign sessions that are individually harmless, dangerous only once collected. Rather than searching the haystack straw-by-straw (i.e. per-session inspection), Magnet does what its name implies: it attracts the relevant needles out of the hay, across sessions and across time, into a compact evidence bundle a detector can act on.
Cognitive AI seeks to move beyond language generation and autonomous task execution toward systems capable of sustained reasoning, adaptive behavior, persistent memory, and self-regulation. While generative and agentic AI have demonstrated impressive capabilities across a wide range of tasks, many fundamental cognitive functions remain fragmented or weakly developed, limiting reliable operation over extended time horizons. This paper presents a taxonomy-driven survey of the major cognitive capability gaps that continue to constrain the development of Cognitive AI. The literature is organized around five dimensions: persistent state modeling, goal-directed autonomy, self-monitoring and control, environment interaction, and learning and adaptation. For each dimension, we review recent advances, identify recurring limitations, and discuss open research challenges. Building on these insights, we outline a conceptual Adaptive Cognitive Intelligence Architecture (ACIA) and examine emerging directions in cognition-centric evaluation. The proposed taxonomy provides a unified framework for organizing existing research, identifying unresolved challenges, and guiding the design of future cognitively capable systems. Together, the taxonomy, architectural perspective, and evaluation framework offer a roadmap for advancing AI systems that exhibit more reliable long-term reasoning, adaptive decision-making, and continual learning. The survey highlights key research opportunities toward more adaptive, reliable, and cognitively capable AI systems, providing a foundation for future progress toward Cognitive AI and, ultimately, Artificial General Intelligence (AGI).
The efficiency of a datacenter rests on its control plane policies. Designing these policies is increasingly hard: the hardware-software stack grows fast, the design space is vast and interdependent, and prototyping a single policy takes months. Agentic AI promises to automate this search. Off the shelf, however, it falls short on three fronts. It is not formal: with no structured, searchable statement of the problem, the search has little structure to exploit and hard constraints are not guaranteed. It is not transferable: each task is solved from scratch, so nothing learned on one task carries to the next. Finally, it is not systematic: relying on the LLM as the sole source of candidates, it explores a narrow slice of the design space and settles into local optima. We introduce AtumAI, a framework that generates datacenter control-plane policies with agentic AI, making the process formal, transferable, and systematic. From a goal stated in plain language, AtumAI autonomously proposes, tests, and refines candidate policies until one satisfies the request. It does so through two components. The Datacenter Task Compiler automates problem formulation: it compiles the request into a formal, machine-checkable, and searchable specification of the task's objectives, constraints, decision variables, and evaluation methodology. The Evolutionary Design Discovery Loop then searches this specification, expanding the search beyond the LLM itself via a diffusion model, an evolutionary algorithm, and a surrogate model. Together, they reduce onboarding a new task from months of engineering to writing its description. We evaluate AtumAI on three control-plane tasks with distinct problem scopes, design spaces, and trade-offs: workload placement, resource scaling, and power management. Across all tasks, the policies generated by AtumAI consistently outperform expert-engineered baselines.
Cost-Effective Automated Judging of Natural-Language Mathematical Proofs
成本效益高的自然语言数学证明自动评判
Grayzel, Benjamin
Abstract
Grading natural-language mathematical proofs is a recurring cost in evaluating math-reasoning systems, and frontier LLM judges are expensive. We ask whether cheap open-weight models can serve as reliable judges given a candidate proof, a ground-truth proof, and a human-grading rubric. On a 200-instance validation sample of IMO-GradingBench, three cheap judges (GPT-OSS 120B, DeepSeek-V4 Flash, Gemma-4 31B) agree with human pass/fail decisions at rates statistically indistinguishable from Claude Opus 4.7 and Gemini 3.1 Pro, at up to $100\times$ lower cost. We had expected a majority vote of the three to be the best budget option; it matched the frontier but did not improve on its strongest member. Extending to the full 1000-instance benchmark and exploring consensus rules, we found that requiring unanimous agreement (all-three-pass) reaches the highest pass-agreement and precision and, on four replicate runs, the smallest run-to-run spread. The headline finding is that cheap judges are competitive with the frontier at one to two orders of magnitude lower cost; as a deployable default we recommend all-three-pass, with the caveat that this rule was identified post-hoc and warrants independent replication.
Chinese Translation
对自然语言数学证明的评分是评估数学推理系统中的一项重复性成本，而前沿的大型语言模型（LLM）评判者则成本高昂。我们探讨了廉价的开放权重模型在给定候选证明、真实证明和人工评分标准的情况下，是否可以作为可靠的评判者。在200个实例的IMO-GradingBench验证样本中，三位廉价评判者（GPT-OSS 120B、DeepSeek-V4 Flash、Gemma-4 31B）与人类的通过/失败决策达成的协议率在统计上与Claude Opus 4.7和Gemini 3.1 Pro无显著差异，且成本低至其的$100 imes$。我们原本预计三者的多数投票将是最佳的预算选项；结果与前沿模型相当，但并未超越其最强成员。扩展到完整的1000实例基准并探索共识规则后，我们发现要求一致同意（全员通过）能够达到最高的通过一致性和精确度，并且在四次重复运行中，运行间差异最小。我们的主要发现是，廉价评判者在成本低一到两个数量级的情况下与前沿模型具有竞争力；作为可部署的默认选项，我们推荐全员通过，但需注意该规则是在事后识别的，值得进行独立验证。
Peer review at major venues is under unprecedented submission pressure, motivating the use of large language models (LLMs) as review assistants. Existing LLM-based reviewers, however, face two structural limitations. First, they map manuscripts directly to reviews, leaving the underlying rubric implicit and entangling its derivation with the judgement. Second, the prevailing paradigms each capture only half of a good review: training-free agents gather broad evidence but produce undirected critiques, while training-based reviewers inherit human discriminative judgement together with its noise and uneven coverage. We introduce RubricReviewer, a fully rubric-driven framework that addresses both limitations. It makes rubric generation an explicit intermediate step, so that both review generation and the final assessment are conditioned on paper-adaptive rubrics. It further combines a training-free agent (Scout) that gathers external evidence with a human-aligned trained model (Aligner) that consumes this evidence, fusing the strengths of both supervision sources. Experiments on real-world submissions show that RubricReviewer produces reviews that are markedly more comprehensive and more discriminative than prior systems, and exhibits the strongest robustness against adversarial prompt-injection attacks. Ablation studies further confirm the necessity of each component.
MemoryForge: Synthesize Lifelong Memory for Human-Like LLM Agents
MemoryForge：为类人大型语言模型代理合成终身记忆
Tang, Bohan, Guo, Yiwen
Abstract
Equipping Large Language Models (LLMs) with human-like personas is crucial for agentic applications, such as role-play and user simulation. Traditional prompt-based methods rely on descriptive conditioning by injecting static textual profiles, which often makes agents show generic behaviors due to a lack of realistic life memory. To fill this gap, we introduce memory-based conditioning, a paradigm inspired by the cognitive psychology, which replaces abstract profiles with an autobiographical memory base, enabling frozen LLMs to dynamically retrieve situation-relevant memory to guide their behaviors. We formalize its enabling task as customized lifelong memory synthesis and propose MemoryForge, a novel framework to synthesize such lifelong memory from brief target personas. MemoryForge has three key components: a context generator for socio-historical grounding, a life organizer for developmental coherence toward the target identity, and a multi-resolution simulator that balances broad temporal summaries with high-fidelity episodic experiences. Experiments on PersonaGym for role-play and SimulatorArena for user-simulation, show that the synthesized memory base by MemoryForge enables frozen LLMs to exhibit more human-like behaviors than strong descriptive conditioning baselines across multiple metrics and LLM backbones.
AgentMemBench: A Systematic Benchmark for Evaluating Long-Term Memory Management Strategies in Conversational AI Agents
AgentMemBench：评估对话式人工智能代理长期记忆管理策略的系统基准
Cherif, Ahmed
Abstract
Long-term memory remains a critical bottleneck for conversational AI agents, whose finite context windows cannot support coherent recall across thousands of turns. We present AgentMemBench, a unified, reproducible benchmark evaluating five memory management strategies under identical conditions: in-context windowing (ICW), external key-value store (EKV), graph-based episodic memory (GEM), compression-based summarisation (CBS), and web-augmented memory (WAM). All are assessed across three public datasets covering long-term multi-session dialogue (LoCoMo), task-oriented document grounding (MultiDoc2Dial), and persona-grounded multi-session chat (MSC), using Recall@k, MRR, nDCG@k, Answer F1, an LLM-judge Faithfulness score, Memory Footprint, and Latency over 491 annotated question turns. Generation and judging both use Qwen2.5-7B-Instruct (4-bit), with greedy decoding for determinism. Our results show that (1) EKV dominates on every quality axis (macro Recall@5 0.792, MRR 0.677, F1 0.156, Faithfulness 0.354); (2) long-range recall is decisive: on LoCoMo, where the gold turn lies many sessions back, ICW, WAM, GEM, and CBS retrieve almost nothing (Recall@5 <= 0.005) while EKV alone reaches 0.573, showing that recency windows, summaries, and entity graphs collapse at long horizons and only dense retrieval scales; (3) CBS is the runner-up on retrieval (0.556); (4) WAM equals ICW on in-corpus recall by construction, since external results carry no in-corpus provenance; and (5) EKV's recall advantage carries a footprint cost (~5,100 vs ~300 tokens for ICW/WAM), an explicit accuracy-efficiency trade-off. We additionally evaluate two published memory systems (MemGPT/Letta, HippoRAG) against the same harness, and release all code, environment, and result artefacts for full reproducibility.
Current text-to-speech systems face a trade-off: autoregres- sive codec language models produce highly intelligible speech but require large-scale models and training data and decode tokens sequentially, while non-autoregressive approaches im- prove speed at the cost of linguistic accuracy. We present DLLM-TTS, a framework that formulates TTS as conditional block discrete diffusion over X-Codec2 neural audio codec to- kens. The model decomposes sequences into blocks and applies masked diffusion within each block while processing blocks se- quentially, learning both local acoustic coherence and global text-speech alignment. During inference, parallel token pre- diction within blocks enables efficient generation with a real- time factor (RTF) of 0.15. A 0.6B-parameter model trained on 20K hours achieves competitive performance on the Seed- TTS-eval benchmark, demonstrating that block discrete diffu- sion language models enable practical and data-efficient speech synthesis with parallel generation.
Multimodal Large Language Models (MLLMs) are increasingly used to interpret Earth observation data, yet their capability to support real-world disaster emergency response remains insufficiently evaluated. Existing remote sensing benchmarks largely rely on static, post-hoc, and expert-processed products, such as gridded reanalysis data, which are difficult to align with operational disaster scenarios where hazards evolve rapidly and decisions must be made under strict time constraints. To bridge this gap, we introduce Obshazard-bench, a real-time, observation-driven benchmark for evaluating disaster intelligence in MLLMs. Unlike image-centric or post-event benchmarks, Obshazard-bench directly integrates raw, high-frequency satellite sounding streams from diverse satellite sensors with concurrent ground-station observations, historical disaster records, and socio-economic indicators, bypassing delayed expert-processing and physical-inversion pipelines. The benchmark covers 8 major disaster categories and 28 sub-categories across more than 60 countries, incorporating over 120 historically documented extreme-event cases and thousands of lifecycle-oriented VQA samples. Moreover, Obshazard-bench further defines a three-stage evaluation taxonomy aligned with the operational disaster workflow: Predictive Crisis Anticipation for pre-disaster risk detection and early forecasting, Active Evolution Reasoning for in-situ disaster tracking and termination prediction, and Multi-faceted Impact Quantification for post-disaster magnitude deduction, humanitarian burden estimation, and socio-economic impact assessment. Experiments on representative general-purpose and Earth-focused foundation models reveal substantial limitations in transforming raw multi-channel physical observations into temporally grounded and decision-relevant disaster reasoning.
Choosing the right large language model (LLM) backbone is the most consequential decision when building a vision-language model (VLM), yet it remains fundamentally unprincipled: compute-based scaling laws fail to generalize across model families, and no framework exists for directly predicting VLM performance before training begins. We propose the Capability-Driven Multimodal Scaling Law, the first cross-family framework that predicts VLM benchmark accuracy from directly observable textual capability. Given a low-dimensional capability score $S$ extracted from LLM textual benchmarks via PCA, we model VLM performance as a function of $S$, with a per-backbone transfer rate and an absorption rate that quantifies data-scaling efficiency. To fit and validate the framework, we train over 150 VLMs on 34 LLMs spanning 7 model families under a strictly controlled recipe. Evaluations on more than 200 textual and 50 multimodal benchmarks show that the law accurately extrapolates transfer rate from models up to 8B parameters to 72B-scale backbones, predicts full VLM training trajectories with high fidelity, and generalizes to entirely held-out model families. Beyond the scaling law, our analysis surfaces actionable insights: certain textual benchmarks negatively correlate with multimodal performance, exposing latent benchmark-gaming behavior; base LLMs outperform instruction-tuned counterparts as VLM backbones due to higher absorption rates and lower data-scaling decay; and different model families occupy distinct positions in the transfer--absorption space. The framework turns backbone selection from costly empirical sweeps into a principled, quantitative decision. Code and data are available at https://github.com/wangq-dev/CDMScaling.
Social simulations built from language-model agents need role-conditioned behavior that can be checked before agents are placed into a simulated population. We introduce an activation-steering screening workflow for role-conditioned agents: define a role profile, extract a role-specific direction, sweep four steering coefficients, evaluate role-profile alignment, and pass or flag each candidate configuration. On OLMo-3-7B-Instruct, we apply the workflow to a mixed 275-role inventory with 228 role-agnostic questions, GPT-4.1-mini prompted role references, and GPT-4.1-mini judges. Role-specific directions receive higher judged role-profile alignment than an assistant-axis directional control from prior persona-vector work, with mean overall scores of 63.2 versus 41.1 across the tested grid. They also preserve high lexical diversity, while the control drops sharply at larger coefficients. The role-level screen is the main practical output: most roles improve as steering increases, but 38 roles decline across all six measured dimensions, showing why simulation builders should choose coefficients per role rather than deploy a uniform high-strength setting. We make our code and evaluation artifacts available at https://anonymous.4open.science/r/anonymous-research-code-5F03/.
Although diffusion models have revolutionized continuous domains like image synthesis through high quality generations and controllable guidance mechanisms, bringing this controllability to the discrete, sequential nature of text remains an open challenge. Meanwhile, current sampling strategies and guidance methods adjust token likelihoods without capturing the broader semantic landscape, leading to a suboptimal balance between fidelity and diversity. In this work, we introduce a novel training-free Semantic-Aware Kernel Entropy (SAKE) guidance method. Our method computes the order-2 R\'enyi entropy over a kernel Gram matrix that captures both cross-token semantic interactions and relative token positions. By linearizing this objective in the embedding space, we derive a tractable guidance signal that dynamically adjusts the sampling distribution, flattening it to encourage exploration during redundancy and sharpening it for fidelity when diverse. Empirical experiments demonstrate that our approach achieves a superior Pareto frontier between fidelity and diversity, and improves multi-sample performance on reasoning-intensive tasks, such as code and mathematics generation, compared to temperature scaling and discrete guidance baselines.
Specialised retrieval agents typically surface higher quality results than general-purpose search, but selecting the optimal agent for a given query remains an open problem. Current approaches route queries based on inferred topic or intent, however intent-based selection is fundamentally limited: it does not incorporate signal from retrieved content, and cannot detect when a topically aligned agent produces low-relevance results. We address this by training a small language model via supervised fine-tuning followed by reinforcement learning to jointly perform agent selection and structured parameter generation for downstream tool calls, using a hierarchical reward function grounded in retrieval relevance along with query-agent topic alignment. This enables the model to learn task-dependent agent suitability from retrieval performance: which agents reliably yield high-relevance results for which query distributions, and when to redirect queries away from specialised agents despite surface-level topical overlap. On a targeted subset of such agent-query mismatches, the trained model achieves an NDCG@10 of 0.918 compared to 0.539 and 0.490 for two LLM baselines (Amazon Nova Lite and Claude Haiku 4.5) that route on intent alone. Overall, it achieves a mean NDCG@10 of 0.771 (+0.177 over Nova Lite, +0.219 over Haiku) with a mean selection latency of 120.1ms, an 82.4% reduction over Nova Lite.
Chinese Translation
专门的检索代理通常能够提供比通用搜索更高质量的结果，但为给定查询选择最佳代理仍然是一个未解决的问题。目前的方法基于推断的主题或意图来路由查询，然而基于意图的选择在根本上是有限的：它不考虑检索内容的信号，并且无法检测到当一个主题对齐的代理产生低相关性结果时的情况。我们通过训练一个小型语言模型，采用监督微调后再进行强化学习，来共同执行代理选择和结构化参数生成，以便进行下游工具调用，使用基于检索相关性的层次奖励函数以及查询-代理主题对齐。这使得模型能够从检索性能中学习任务依赖的代理适用性：哪些代理在特定查询分布下可靠地产生高相关性结果，以及何时尽管表面上存在主题重叠也要将查询重定向到其他代理。在这样的一组目标代理-查询不匹配中，训练后的模型在 NDCG@10 上达到了 0.918，而仅基于意图路由的两个大语言模型基线（Amazon Nova Lite 和 Claude Haiku 4.5）的 NDCG@10 分别为 0.539 和 0.490。总体而言，该模型的平均 NDCG@10 达到了 0.771（比 Nova Lite 提高了 0.177，比 Haiku 提高了 0.219），平均选择延迟为 120.1 毫秒，比 Nova Lite 减少了 82.4%。
Real-world document tasks often ask professionals to answer questions from annual reports, regulations, clinical guidelines, and technical manuals that span hundreds or thousands of pages. Some questions also require comparing related reports. Reliable long-document understanding is therefore a prerequisite for using LLMs in compliance, clinical, financial, and engineering workflows, where decisions must be traceable to specific evidence pages and the cost of an unsupported answer is high -- yet most existing benchmarks still measure short-context or single-page QA. We introduce XL-DocBench, a fully human-verified benchmark for extra-long document understanding, with 1,519 retained questions from six professional domains and contexts up to 2,303 pages. XL-DocBench goes beyond page-level lookup. 1,103 examples (72.6\%) use multiple evidence pages. The final set also includes 556 questions (36.6\%) that use tables, charts, or figures, and 165 questions (10.9\%) that require evidence from multiple documents. Each question has one of twelve reasoning labels, expert-annotated evidence pages, a typed verification rule, and an answer format, including 218 None-answer cases. We build the benchmark with a tree-guided synthesis pipeline followed by artifact filters and full verification by 194 human experts. By coupling extra-long professional contexts with page-level evidence and typed rules, XL-DocBench fills a gap left by prior single-page, short multi-page, or text-only long-context benchmarks, and lets future work attribute system failures to retrieval, evidence use, or rule following rather than to a single leaderboard score. The results show that current systems still struggle with long contexts, multi-page evidence, and structured reasoning over professional documents.
Trustworthiness Costs of Domain Adaptation in Small Language Models:A Cross-Architecture Empirical Study
小语言模型领域适应的可信度成本：跨架构实证研究
Paramkusham, Ramesh B.
Abstract
Domain adaptation of small language models (SLMs) has emerged as a practical strategy for deploying capable NLP systems in resource-constrained, high-stakes environments including healthcare, legal services, and financial analysis. While performance gains from parameter-efficient fine-tuning are well characterised, the corresponding impact on trustworthiness (factual calibration and adversarial robustness) remains poorly understood. This paper presents the first systematic cross-domain, cross-architecture empirical study quantifying the trustworthiness cost of domain adaptation across three SLM architectures (TinyLlama 1B, Gemma-2 2B, Llama 3.2 1B), three domains (healthcare, legal, finance), two training-data conditions (benign and adversarially perturbed), and four fine-tuning strategies (baseline LoRA, Safety-DPO, Dark Experience Replay, and Task Arithmetic LoRA, TA-LoRA). Trustworthiness is evaluated through TruthfulQA MC2 (factual calibration) and HarmBench ASR (adversarial robustness) across all 216 experimental configurations with three random seeds. Three principal findings emerge. First, baseline QLoRA domain adaptation produces minimal TruthfulQA MC2 change across all model-domain combinations (mean |Delta TQA| < 0.02). Second, adversarially perturbed training data consistently improves domain adaptation quality (Delta loss approximately -0.040) without worsening trustworthiness benchmarks. Third, none of the three safety-preserving strategies reduced adversarial harm susceptibility: Safety-DPO was effectively neutral (mean Delta ASR < 0.001), while Dark ER and TA-LoRA increased mean HarmBench ASR by +0.171 and +0.155 respectively in safety-aligned models (Gemma-2 2B, Llama 3.2 1B), with individual configurations exceeding +0.45. These results challenge the assumption that replay-based and arithmetic-merge strategies transfer alignment to domain-adapted SLMs.
Predicting Startup Exit from Textual Descriptors - A Computational Linguistics Framework
从文本描述预测创业公司退出 - 一种计算语言学框架
Saruggia, Alberto M. G., Germano, Sebastien
Abstract
This study shows that textual descriptors alone can predict early-stage startup success, defined as Exit, without relying on contextual, financial, or human capital variables. Using venture capital-curated datasets covering 7,419 startups over 20 years, the research isolates text-based framing variables and engineers 850 features through startup narrative mapping. Data subsets and vector embeddings are evaluated for statistical significance, followed by supervised machine learning experiments across six models. LightGBM achieved the highest predictive performance (F1 = 0.48), while textual descriptors alone achieved F1 = 0.30, confirming the standalone predictive value of founder narratives. Feature analysis shows that optimized densities of hyping markers, including adjectives, jargon, and buzzwords, are associated with higher Exit probability, whereas excessive statement or name length reduces it. The study also introduces a quantifiable Hyping Score for venture capital applications, demonstrating that startup framing provides measurable signals for predicting Exit under conditions of high information asymmetry.
The success of connectome mapping now shifts the challenge of understanding the nervous system to the interpretation of neural circuits. Here, we devise a new automated method, LLantia (LLM automated neural circuit inference and analysis), to systematically infer neural circuit function and the role of its component neural cell types. Our approach distills descriptions of cell type function from the literature and, in combination with the connectome, then infers the function for all other cell types, which serves as a basis for subsequent neural circuit function inference. Results are structured hierarchically, with different possible circuit functions organised under multiple possible behavioural and physiological contexts, and each circuit function composed of subcircuit descriptions alongside relevant cell types to facilitate both backtracking to known, published information and support further experimental research. We illustrate our method by inferring cell type function for all cell types of the adult fruit fly brain and for select broader circuits within, and validate our findings, including by cross-checking with literature published after the release date of our analysis.
LLM-native advertising embeds sponsored content directly into model-generated responses, shifting the unit of sale from a fixed slot to a moment within an evolving conversation. Existing LLM ad-auction mechanisms primarily operate within a single response, settling the winner but not the timing. The extension is nontrivial: with one native insertion opportunity per session, the stopping time depends on bids, coupling timing with allocation, so static truthfulness arguments no longer apply. We propose the LLM-based Optimal Stopping Dynamic Auction (LLM-OSDA), a dynamic cost-per-click auction that integrates Bellman optimal stopping, winner allocation, and envelope pricing. A bid-independent LLM layer estimates contextual click quality and seamlessly renders the winning ad, while bids enter only the committed auction mechanism. Under an exact Bellman oracle, the expected discounted-click allocation is monotone in each advertiser's bid, and the corresponding envelope payment makes truthful bidding weakly dominant in expectation. For practical deployment, a learned StopNet approximates the Bellman action values. We show that its decisions differ from the optimal policy only near the stopping boundary and bound the resulting incentive loss in terms of its approximation error. Experiments on a simulated conversational advertising corpus show that LLM-OSDA improves net revenue by 11 percent over the strongest fixed-timing baseline while maintaining comparable user retention. Code is at https://github.com/2025Fang2025/llm-osda.
We introduce DiffusionGemma, an experimental open-weight language model that uses discrete diffusion to generate text at exceptionally high speed. Rather than decoding one token at a time, DiffusionGemma iteratively refines blocks of 256 tokens in parallel, avoiding the sequential decoding bottleneck of conventional autoregressive (AR) large language models. Instead of training from scratch, we obtain DiffusionGemma by fine-tuning the mixture-of-experts Gemma 4 model with 3.8B activated and 25.2B total parameters. Our compute-efficient two-stage training pipeline uses fewer than 10% of the starting AR model's total training token budget. The first stage uses supervised fine-tuning to teach bidirectional denoising, while the second stage combines reinforcement learning with sampler distillation to jointly improve generation quality and inference efficiency. DiffusionGemma establishes a new Pareto frontier for the trade-off between generation speed and model capability. Averaged across our full evaluation suite, it generates around 20 tokens per forward pass and achieves roughly 1,500 output tokens per second on a single NVIDIA H100 GPU, which is substantially faster than AR models even with state-of-the-art speculative decoding. DiffusionGemma also retains the starting model's support for thinking mode, multimodal inputs, and long contexts. Despite diffusion fine-tuning, it remains capable of AR generation with only minor performance degradation, suggesting a path toward hybrid diffusion-AR decoding.
A Constitution-Grid Instrument for Data-Efficient RL Alignment (C-Guard)
一种用于数据高效强化学习对齐的宪法网格工具 (C-Guard)
Zhang, Lily
Abstract
Conflicting objectives are general in RL alignment, and training on them data-efficiently is hard. Training a safety guard with RL means optimizing two objectives that conflict: catch real harm, and do not refuse benign prompts. Our finding is that over-refusal improves 22.4% to 12.8%, while under-refusal on adversarial attacks silently worsens 0.27 to 0.33. We present C-Guard, a constitution-grid instrument that generates the RL training data, and C-LIM, a per-cell learnability score that decides each cell's move: prune, densify, amend, expand. C-LIM flags the dead-weight data region before any training budget is spent: 187 untargeted rows had bought zero gain, and our method lifts the same region's learning impact 0.733 to 0.80. Code and the constitution are open-sourced.
Averaging Bias: Human Faithfulness Annotations are not Locally Faithful
平均偏差：人类信度标注并不局部忠实
Zhang, Huajian, Feng, Yiyang, Zhou, Jiawei
Abstract
Evaluation of faithfulness of text summarization treats a model generated summary as faithful only if every of its sentences is supported by the source document: a strict conjunctive rule under which a single unsupported sentence makes the whole summary unfaithful. Yet most faithfulness benchmarks collect only one global human annotation label per summary. We ask whether such global human labels actually implement the conjunctive rule. We hypothesize that annotators may accept a summary as faithful when most sentences are faithful, not only when all are faithful. To test our hypothesis, we use five large language model (LLM) judges as per-sentence raters across four widely used faithfulness benchmarks. We find that global human labels correlate better with the average of per-sentence LLM judgments than with the implementation of the strict conjunctive rule. A manual review confirms that a substantial fraction of summaries labeled faithful by humans contain genuine local factual errors. We call this tendency Averaging Bias. Our results reveal that human labels on widely used faithfulness benchmarks contain measurable Averaging Bias, calling for carefully structured designs for trustworthy human annotations
Large Language Models (LLMs) perform strongly in English medical tasks but degrade substantially in Arabic, a gap widely attributed to limited training data. We systematically investigate this assumption via tuned lens probing and causal activation patching, and find that Arabic medical knowledge is present in intermediate model representations but fails to surface at the output. This mechanistic insight motivates a targeted adaptation strategy: rather than fine-tuning the full network, we propose Targeted Low-Rank Adaptation (TLoRA), restricted to the layer window where cross-lingual representations diverge, upstream of the output layers where the failure manifests. We evaluate TLoRA on multiple-choice medical QA, where our approach outperforms full-network LoRA, zero-shot, and few-shot baselines. We further evaluate it on short-answer generation and multi-turn clinical dialogue, where it performs competitively without the need for task-specific finetuning. We additionally introduce AraClinicDialog, a clinician-constructed Arabic medical dialogue benchmark in MSA with validated variants across four Arabic dialects. Together, these contributions demonstrate that mechanistic diagnosis can serve as a practical guide for targeted adaptation in underrepresented-language medical LLMs.
Agentic LLMs exhibit three consequential tool-use failures: invalid arguments (validity), unnecessary calls (over-calling), and omitted calls when tools are needed (missing). We find that a small, failure-specific set of MLP neurons could distinguish such failures with linearly separable decision boundaries. Building on this observation, we introduce PRISMS (Probing Representations In Support of Monitoring and Steering), a closed-loop framework that shares a failure-specific neuron basis between sparse detection and activation steering. PRISMS selects contribution-critical MLP neurons and fits an L1-regularized detector on their activations. Across six models from the Qwen3, Llama, and Gemma families, over-calling and missing are detected at the pre-generation prompt boundary with ROC-AUC 0.90-1.00, while validity is detected from the generated tool-call span with ROC-AUC 0.86-0.90. These results are achieved with highly sparse readouts: only 1-2 MLP neurons for missing, 2-16 for over-calling, and approximately 128 for validity. These sparse detectors match or outperform dense residual-stream baselines using 23-627 times fewer features. The shared neuron basis also supports bidirectional control over tool-calling behavior, suppressing unnecessary calls and eliciting omitted ones. PRISMS therefore gates intervention on predicted failure risk to mitigate the collateral effects of unconditional steering. Across all six models, PRISMS reduces pooled over-calling rate by 80% (from 0.131 to 0.026) while increasing tool-required accuracy by 14.2 percentage points (from 0.689 to 0.831). PRISMS thus provides lightweight failure detection and selective intervention across model families.
Do frontier vision-language models present a coherent Theory-of-Mind (ToM) profile across tasks, matching the same human reference group, or does that profile fragment from one paradigm to the next? We evaluate a shared panel of nine frontier VLMs on two psychology-derived benchmarks: the Keysar Director Task (visual perspective-taking under egocentric interference) and the Frith-Happ\'e animated triangles scored with the Castelli rubric (intention attribution from pure motion). On the Director Task, without chain-of-thought, the panel makes the egocentric error on 78\% of trials like children rather than adults; variation is substantial across models, and reasoning rescues several models. On the triangles, the panel under-attributes intention: its ToM profile sits more than three times closer to the high-functioning-autistic-adult (HF-ASD) mean than to the typical-development-adult (TD) mean, while Goal-Directed and Random stay near TD. No model is nearest TD on both tasks; the model that looks adult-like on the Director Task falls on the HF-ASD side on the triangles, and the most TD-like model on the triangles is child-like on the Director Task. We report group-level descriptions, not diagnostic labels for any model.
Sixteen language models drawn from ten families produced, on average, the semantic diversity of 1.69 distinct formulations of a psychotherapeutic case, against a single-model baseline of 1.43 from one model's own runs. Ensembles place more than one reading before a decision-maker on the premise that several models supply several perspectives. Dispersion over their outputs is measured both as diversity and as uncertainty, and both traditions validate it against a correctness criterion that this task does not admit. Measuring diversity is a solved problem: the Vendi Score, the exponential of the von Neumann entropy of a similarity matrix, is an effective number of distinct elements. What a single aggregate does not say is where the diversity comes from. We define a per-model dissent contribution, the complement of a model's mean similarity to the other members of its ensemble: a magnitude from the same matrix, not a decomposition of the spectral index, whose maximum identifies the most divergent voice. Crossing model and case, we test as a preregistered hypothesis whether model identity accounts for a non-zero share of the variance in dissent, and characterise the structure that test detects. The panel formulated fifteen stratified vignettes, yielding 7,082 formulations for analysis. Model identity was a detectable structuring factor of the dissent that remained, but the usual categories recovered it only partly: scale differences pointed in opposite directions across pairs, family grouped models on only five two-member lines, and the most divergent voice changed with panel composition, so that the surfaced outlier describes the ensemble rather than the model. Dissent did not track the interpretive openness for which the case bank was stratified; it was organised by clinical content instead, leaving the dispersion an ensemble produces a property to measure rather than assume.
Comparing and Modeling Argumentation in German Political Communication across Arenas
比较与建模德国政治传播中的论证在不同场域中的表现
Vikhrova, Nina, Kühling, Johannes, Haunss, Sebastian, Padó, Sebastian
Abstract
Deliberation, involving the formulation and exchange of arguments, forms an integral part of political decision making in democracies. Argumentation patterns however differ substantially across different political arenas, such as plenary speeches and committee meetings. However, despite a lot of interest in argumentation, there is comparatively little computational work on analyzing differences in patterns of political argumentation between arenas. Our work addresses this research gap. First, we present a 17k-sentence corpus with annotation for argumentative passages (argument and their justifications, both their boundaries and their categories) across three German political arenas (plenary speeches, committee meetings, and press conferences), keeping the topic (COVID-19) constant. Our analysis of the corpus finds that contrary to expectations, justification by domain-specific expertise is more frequent in press conferences than in committee meetings. Second, we present a pilot study on automatically identifying such argumentative passages. The results show that boundaries are hard to pin down, and models predictions additionally suffer from confirmation bias.
SeDeM: Selective Decompression of Hidden-State Memories for Long-Context Question Answering
SeDeM：用于长上下文问答的隐状态记忆选择性解压缩
Haghifam, Maryam, Cong, Jason, Sun, Yizhou
Abstract
Long-context inference with large language models (LLMs) is costly: self-attention during prefill scales quadratically with sequence length, and the key-value (KV) cache grows with the number of processed tokens. Larger context windows also do not ensure reliable evidence use. Context compression reduces this cost, but many soft-compression methods use LLMs as compressors and rely on compact memory tokens both to preserve information and to condition the decoder. We propose SeDeM, a selective decompression framework that decouples compact memory storage from decoder conditioning. An LLM extracts hidden states from a chosen intermediate Transformer layer, a lightweight compressor stores them as memory blocks, a query-conditioned selector selects relevant blocks, and a decompressor expands only the selected blocks into hidden states compatible with an intermediate decoder layer. Thus, the decoder avoids both full-context processing and direct generation from highly compressed memory slots. On four long-context QA benchmarks, SeDeM achieves higher QA scores than the evaluated compression baselines in both 1B and 3B same-backbone settings, and with the 3B backbone exceeds full-context fine-tuning on three datasets. The learned selector uses block-level evidence supervision during training. SeDeM also reduces online time-to-first-token and improves autoregressive decoding throughput relative to ICAE.
Progress in large language models is often summarized using a single scalar measure, such as a time horizon, a latent ability estimate, or an aggregate benchmark score. These summaries capture the overall performance, but they do not test whether progress is distributed differently across task difficulty. We find that most of the apparent shift in gains toward harder tasks does not reflect a change in the shape of the difficulty-response curve. On METR time-horizon data, a single Rasch model with rising ability reproduces this pattern, so it is largely explained by ceiling effects rather than a qualitative change in capability. This echoes how the choice of metric can make claimed emergent abilities look like a property of the models themselves. We then identify a smaller hard-task effect that survives this control. Isolating it is difficult on agentic benchmarks, because newer models are usually run with newer agentic harnesses, so a gain on hard tasks cannot be assigned to the model or its scaffold. We break the confound with LiveCodeBench, a public competitive programming benchmark that runs no agentic scaffold while pairing dated models with an exogenous difficulty ordering. After accounting for the rise in overall ability, models released after September 2024 still gain on the hardest problems beyond what their easy and medium performance predicts, by about +0.40 logits under our most conservative assumption, raising the hard-problem solve rate from roughly 18% to 25%. The effect is led by the strongest reasoning models and holds for hard tasks that need only short reasoning, not autonomy over long horizons. We present this as a result specific to competitive programming, since our clean identification rests on a single coding benchmark. We release the LiveCodeBench Difficulty Panel (66 dated models x 1,055 problems) and our analysis code.
Chinese Translation
大型语言模型的进展通常使用单一的标量指标进行总结，例如时间范围、潜在能力估计或综合基准分数。这些总结捕捉了整体性能，但并未测试进展是否在任务难度上分布不同。我们发现，向更难任务的收益明显转变大部分并不反映难度-响应曲线形状的变化。在METR时间范围数据上，单一的Rasch模型（Rasch model）具有上升的能力，重现了这一模式，因此它在很大程度上是由天花板效应（ceiling effects）解释的，而非能力的定性变化。这与度量选择如何使声称的突现能力看起来像是模型本身的属性相呼应。随后，我们识别出一个较小的难任务效应，该效应在此控制下仍然存在。在代理基准上隔离它是困难的，因为较新的模型通常与较新的代理框架一起运行，因此在难任务上的收益无法归因于模型或其支架。我们通过LiveCodeBench打破了这种混淆，这是一个公共的竞争编程基准，不运行任何代理支架，同时将过时的模型与外生的难度排序配对。在考虑整体能力上升后，2024年9月之后发布的模型在最难问题上的收益仍超出其在简单和中等任务上的表现预测，约为+0.40 logits（在我们最保守的假设下），将难问题的解决率从大约18%提高到25%。这一效应主要由最强的推理模型引领，并且适用于仅需短期推理而非长时间自主性的难任务。我们将此视为竞争编程特有的结果，因为我们的清晰识别基于单一的编码基准。我们发布了LiveCodeBench难度面板（66个过时模型 x 1,055个问题）及我们的分析代码。
Deep research agents are often trained on expensive, environment-grounded tool-use trajectories that require repeated retrieval, document inspection, and report evaluation. We introduce Deep Research Pretraining (DRP), an offline framework that derives predictive navigation supervision from naturally occurring evidence structures. Given a citation-bearing or hyperlinked passage, DRP constructs a proxy research objective, recovers linked evidence and graph-related alternatives, and converts them into search-open-write trajectories. This teaches models what to search for, which documents to inspect, and how to synthesize evidence, without a live retrieval environment or executed policy rollout. We instantiate DRP on scholarly citation graphs (DRP-Paper) and Wikipedia hyperlinks (DRP-Web), continually pretrain separate Qwen3-14B-Base models on 1B tokens, and fine-tune them on controlled fractions of 13K agent trajectories. Across five independently sampled subsets at each low-data budget, both variants consistently outperform matched no-DRP models on DeepResearch Bench. With one quarter of the SFT data, DRP-Web even surpasses a fixed no-DRP full-data checkpoint, with gains transferring to ResearchQA, WebWalkerQA, and SimpleQA. Starting from matched low-data SFT checkpoints, the DRP-Web advantage also persists through subsequent agentic RL. Source-matched and evidence-mismatch controls indicate that these improvements arise from evidence-conditioned navigation rather than domain exposure or agent-format imitation. DRP thus provides a promising complementary approach to trajectory-based agent training.
Chinese Translation
深度研究代理通常在昂贵的、基于环境的工具使用轨迹上进行训练，这些轨迹需要重复检索、文档检查和报告评估。我们提出了深度研究预训练（Deep Research Pretraining, DRP），这是一种离线框架，从自然发生的证据结构中推导出预测导航监督。给定一个带有引用或超链接的段落，DRP构建一个代理研究目标，恢复链接的证据和图相关的替代方案，并将其转换为搜索-开放-写入轨迹。这教会模型搜索什么、检查哪些文档以及如何合成证据，而不需要实时检索环境或执行策略展开。我们在学术引用图（DRP-Paper）和维基百科超链接（DRP-Web）上实例化DRP，持续在10亿个标记上对独立的Qwen3-14B-Base模型进行预训练，并在控制的13K代理轨迹的部分上进行微调。在每个低数据预算下的五个独立抽样子集上，这两种变体在DeepResearch Bench上始终优于匹配的无DRP模型。使用四分之一的SFT数据，DRP-Web甚至超过了固定的无DRP全数据检查点，其增益转移到ResearchQA、WebWalkerQA和SimpleQA。从匹配的低数据SFT检查点开始，DRP-Web的优势在随后的代理强化学习中也持续存在。源匹配和证据不匹配的控制表明，这些改进源于证据条件导航，而不是领域暴露或代理格式模仿。因此，DRP为基于轨迹的代理训练提供了一种有前景的补充方法。
Multi-Token Prediction (MTP) has emerged as an effective paradigm that augments a shared Large Language Model backbone with auxiliary heads, training the model to predict several future tokens in parallel to enrich its supervision signal and accelerate inference. However, existing training frameworks adopt a rigid, fixed-length prediction horizon, disregarding the highly non-uniform information density of natural language and code. Forcing the auxiliary heads to predict across high-entropy semantic boundaries injects noisy, conflicting training signals; because these heads share the backbone's latent representations, the resulting gradients backpropagate and interfere with the model's core capabilities. We propose AdaMTP, an adaptive training paradigm that dynamically aligns the prediction horizon with the intrinsic predictability of the sequence. At its core, an entropy-based segmentation algorithm leverages the base model to detect sudden surges in uncertainty as semantic boundaries, partitioning sequences into variable-length groups. Each token is assigned an adaptive prediction depth, and a dynamically masked MTP objective suppresses the loss for predictions that cross these boundaries, attenuating the noisy gradients that degrade the backbone. Across mathematical reasoning, code generation, and general benchmarks on three backbones (Llama-3.1-8B, Qwen-2.5-7B, Gemma-3-12B), AdaMTP consistently outperforms standard MTP in both task performance and inference speedup.
Chinese Translation
多标记预测（Multi-Token Prediction, MTP）作为一种有效的范式，通过辅助头增强共享的大型语言模型（Large Language Model）主干，训练模型并行预测多个未来标记，以丰富其监督信号并加速推理。然而，现有的训练框架采用刚性、固定长度的预测视野，忽视了自然语言和代码中信息密度的高度不均匀性。强迫辅助头在高熵语义边界上进行预测会注入噪声和冲突的训练信号；由于这些头共享主干的潜在表示，导致的梯度反向传播会干扰模型的核心能力。我们提出了AdaMTP，一种自适应训练范式，动态地将预测视野与序列的内在可预测性对齐。在其核心，基于熵的分割算法利用基础模型检测不确定性的突然激增作为语义边界，将序列划分为可变长度的组。每个标记被分配一个自适应的预测深度，而动态掩蔽的MTP目标抑制跨越这些边界的预测损失，从而减弱降级主干的噪声梯度。在数学推理、代码生成以及三个主干（Llama-3.1-8B、Qwen-2.5-7B、Gemma-3-12B）上的一般基准测试中，AdaMTP在任务性能和推理加速方面始终优于标准MTP。
Recent Text-to-SQL systems increasingly rely on multi-turn interaction, execution feedback, and reinforcement learning. However, most existing methods use execution correctness only as a trajectory-level reward, which provides limited guidance for identifying the SQL decisions responsible for success or failure. We propose SERL-SQL, a selective execution-grounded reinforcement learning framework for multi-turn Text-to-SQL agents. SERL-SQL samples on-policy SQL interaction trajectories and uses a training-only teacher to re-score student actions with execution feedback. The resulting teacher--student likelihood gap is converted into bounded, masked weights that reweight GRPO advantages only on SQL and tool-action tokens. In this way, task rewards preserve the optimization direction, while execution hindsight provides localized credit assignment. Experiments on BIRD, Spider, and cross-domain benchmarks show that SERL-SQL achieves competitive performance, reaching 76.56% execution accuracy on BIRD-Dev and 89.92% on Spider-Test. Moreover, our reward-based selection strategy closely approaches the oracle Best-of-N upper bound and consistently outperforms consistency-based selection, showing that SERL-SQL produces high-quality candidates that can be reliably identified by lightweight execution-grounded rewards. Our code will be released at https://github.com/Ffunkytao/SERL-SQL.
The methodology of Constructing the Large-Scale Dataset for Detecting Presuicidal and Anti-Suicidal Signals in Social Media Texts in Russian
构建用于检测社交媒体文本中自杀前和反自杀信号的大规模数据集的方法论（The methodology of Constructing the Large-Scale Dataset for Detecting Presuicidal and Anti-Suicidal Signals in Social Media Texts in Russian）
The suicide is a terrifying act of a person who is misled by his own mental state. This problem arises across many countries. Sadly, Russia also has quite high number of persons who committed suicide. Luckily, a subset of these people writes their struggles in social media, allowing a way to find them and help. However, these valuable texts disappearing in many irrelevant texts which is considerably slowing down the decision process about person's suicidal risk. To tackle this problem, in this work we have presented a detailed methodology of building the dataset for detecting texts that describe presuicidal and anti-suicidal signals. This methodology describes the process of instruction and class table creation, the process of annotation, verification and post-annotation correction. Guiding by this methodology, we collect and annotate a large-scale Russian dataset with more than 50 thousand texts from social media. We provide a count statistic of the dataset as well as common problems in annotation. We also conduct basic experiments of building the classification models to show the on go performance on different levels of annotation. Furthermore, we make the dataset, code and all materials publicly available.
The Learning Objective Governs Perceptual Narrowing: A Cross-Lingual, Layer-Wise, Ten-Seed Study of Self-Supervised Speech Encoders
学习目标主导感知收窄：一项跨语言、逐层、十种种子的自监督语音编码器研究
Yoo, Sejin
Abstract
Perceptual narrowing---the developmental loss of non-native phoneme discrimination in the first year of life \citep{werker1984}---is a canonical developmental finding, yet \emph{what learning objective produces it} remains open. We train a \(\sim\)7\,M-parameter Transformer encoder on child-directed and read speech and evaluate phoneme ABX in English, French, and Mandarin over ten seeds, the seed as the unit of replication. Six results. \textbf{(1)}~The objective sets the direction of cross-lingual transfer: reconstruction (masked mel-prediction) degrades non-native discrimination, prediction (frame-contrastive) improves it---a same-encoder, same-data gap of \(+0.051\) in first-layer Mandarin ABX (\(p=3\times10^{-8}\)), unanimous in sign across twenty runs. \textbf{(2)}~That decline combines a large arm-intrinsic difficulty gradient with a smaller language-specialization effect (matched vs.\ mismatched \(+0.022\), \(p=10^{-4}\), all four layers). \textbf{(3)}~Against a language-symmetric raw-mel floor, reconstruction pushes the first layer \emph{below} the discriminability of its input; prediction pushes it \emph{above}. \textbf{(4)}~Read speech gives a \(3.6\times\) steeper non-native decline than child-directed speech. \textbf{(5)}~The customary three-seed budget cannot see this reliably: an effect unambiguous at ten seeds is called significant by as few as 70\% of three-seed subsets. \textbf{(6)}~Six objective configurations---sharpening, compression, consolidation, their composition, and word-level semantic grounding in two forms---fail to produce the full developmental signature (native improves \emph{and} non-native declines): a single objective moves both languages the same way because it acts on a shared representation. We conclude that the objective, not the architecture, is the first-order determinant of narrowing-shaped representational change.
Rethinking and formalising the state across languages: a unified computational learning theory account
重新思考和形式化跨语言的状态：统一的计算学习理论视角
Idrissi, Mohamed El
Abstract
The linguistic notion of state has traditionally been restricted to the construct (annexation) state of Afroasiatic languages and treated as a language-specific morphosyntactic phenomenon. This article argues instead that the state is a systemic, context-dependent morphosyntactic mechanism that selects grammatical templates across synthetic languages. Within the Template-Based Modular Cognitive framework, taking Riffian as its primary empirical basis, the proposed theory provides a unified explanation for diverse nominal marking patterns traditionally analysed independently and is formalised as a symbolic computational model in which the state is represented by a set-valued function over grammatical templates. A learning algorithm based on finite-set operations acquires and predicts state-dependent grammatical configurations. Beyond nominal morphology, the framework has broader implications for theories of nominal structure and lexical cognition, in particular offering a unified analysis of determiner-noun structure. These results suggest that the state constitutes one instance of a broader class of syntactically conditioned dependencies that also includes agreement and grammatical case.
S$^4$R: Selective Sampling, Subspaces, and Sparse Reconstruction for Compressed Long-Context KV Caching
S$^4$R：用于压缩长上下文 KV 缓存的选择性采样、子空间和稀疏重构
Han, Jialong, Wu, You, Tu, Kewei
Abstract
The growth of context window lengths in Large Language Models (LLMs) significantly enhances their long-context capabilities but incurs prohibitive memory costs due to the Key-Value (KV) cache. Although low-rank compression of KV cache is a promising remedy, existing methods face a dilemma: offline approaches depend on external calibration data, whereas online approaches incur substantial compute for full-prompt decomposition and reconstruction. In this paper, we propose S$^4$R, which builds low-rank subspaces from selectively sampled tokens and computes attention over a sparsely reconstructed KV representation. S$^4$R uses prompt-aware initialization to build initial key/value bases from a representative prompt subset, trading off calibration-data dependence against prefilling cost. Because fully reconstructing the cache at every decoding step is prohibitively expensive and hurts throughput, we further adopt sparse reconstruction to retain only informative positions during decoding. Extensive experiments on LongBench and RULER with Llama and Qwen model families show that S$^4$R achieves up to 5$\times$ KV compression with near full-cache accuracy, combining the efficiency of fixed compression with the adaptability of prompt-dependent methods.
Native Multilingual Chain-of-Thought Reasoning in Low-Resource Southeast Asian Languages
低资源东南亚语言中的本土多语言链式思维推理
Lim, Sean Gip, Tjhi, William Chandra, Chieu, Hai Leong
Abstract
Large Language Models have achieved substantial progress in reasoning capabilities. Yet in low-resource native settings, many suffer from cross-lingual collapse, reverting to English during intermediate steps that require complex logical reasoning. This presents a cold-start bottleneck for policy optimization, whereas standard fine-tuning risks catastrophic forgetting due to cross-lingual representation drift. To address these challenges, we introduce the Onramp-Sequence Cross-Distillation (OSCD), a post-training algorithm that projects high-resource reasoning trajectories into low-resource vocabulary subspaces during generative training rollouts via an integrated translator agentic loop, ensuring the stable and efficient translation of dynamically generated reference samples for fine-tuning. This is coupled with joint-embedding semantic alignment of both reference and target-language reasoning traces, thereby bridging the pairwise cross-lingual representational gaps. Comprehensive evaluations using the AIME25 and HMMT25 benchmarks demonstrate that OSCD yields up to 3.2 times overall improvements in native Southeast Asian languages for mathematical reasoning, of which the joint-embedding semantic alignment component contributes up to 6.4% improvements in linguistic debiasing over translation-only baselines.
DE-NER : Zero-shot Named Entity Recognition via Dialogue Elicitation of Large Language Models
DE-NER：通过对话引导大语言模型实现零样本命名实体识别
Zhang, Xuankang, Liu, Jiangming
Abstract
Recent advancements of zero-shot Named Entity Recognition (NER) establish strong baselines by formulating sequence labeling into question answering where Large Language Models (LLMs) can be naturally adopted. However, existing LLM-based zero-shot NER methods suffer from the limitations of prompt and demonstration engineering. To address these issues with minimal human interventions, we introduce DE-NER, a dialogue elicitation framework which elicits the chatting ability of LLMs to fully extract the knowledge encoded in LLMs. Our experiments demonstrate that the proposed method outperform the competitive baselines in zero-shot settings across multiple benchmarks, with an average improvement of 3.75\% F1 points. Codes are released in https://github.com/kkkenshi/DE-NER.
Loanword or Switch? The Annotation Boundary, Not the Model, Drives Kazakh-Russian Code-Switching Identification
借词还是切换？注释边界，而非模型，驱动哈萨克语-俄语代码切换识别
Savelyev, Bogdan
Abstract
Off-the-shelf LID and letter heuristics over-label Kazakh-Russian social text as mixed: Russian loanwords inside Kazakh look like code-switching under a shared Cyrillic script. We release a document-level gold LID set whose guideline keeps integrated borrowings as Kazakh and reserves mixed for clause-level switches, plus a mixed-only sentiment pool used after LID in a filter-first cascade. On a shared LID test, FastText, Lingua, raw and windowed HeLI, character-trigram NB, and XLM-R range from weak to strong performance. The gap shows the bottleneck is the loanword-vs-switch annotation boundary, not model class alone.
Writing-System-Level Tokenizer Adaptation for Byte-Level BPE
字节级 BPE 的书写系统级标记器适配
Didenko, Bohdan
Abstract
Pretrained byte-level BPE tokenizers can segment underrepresented languages inefficiently. Replacing a tokenizer changes the meaning of nearly every token ID, while vocabulary expansion enlarges the model's embedding and output matrices. We study post-hoc adaptation that keeps the model-vocabulary size fixed and preserves most existing token-to-ID assignments as a construction-time compatibility property. Directly transferring tokens from a language-specific tokenizer does not guarantee derivability through the target BPE merge graph: an inserted entry can conflict with the target's greedy merge ranks. We formalize this failure as the merge ordering problem and introduce BPE-guided insertion, which builds each transferred token through a target-reachable decomposition. Our pipeline uses script-aware row selection to limit collateral fragmentation, reconstructs target-script byte-level prerequisites, and applies guided insertion to maintain merge-graph reachability. On Ukrainian adaptations of Nemotron and GPT-OSS, it reduces token counts by 33.5% and 36.6%, keeps changes on English and the evaluated four-language European aggregate within 0.05%, and retains 78.5%/77.3% of original model-vocabulary rows at the same IDs. Constraint-matched global and frequency-based removal achieve similar Ukrainian compression but increase English/European token counts by 0.7-2.2%; fresh same-size retraining compresses Ukrainian slightly more but retains effectively no same-ID rows and increases English token counts by 7.6-8.6%. The reallocation increases token counts on the evaluated three-language Cyrillic micro-aggregate by 6.7%/10.1%. Structural audits find all 28,134/45,398 inserted BPE nodes reachable under ordinary rank-ordered merging and no retained same-ID model-vocabulary entry newly broken. We release all tokenizers and code.
Chinese Translation
预训练的字节级 BPE 标记器在对低资源语言进行分词时效率较低。替换标记器会改变几乎每个标记 ID 的含义，而词汇扩展则会增大模型的嵌入和输出矩阵。我们研究了一种后期适配方法，该方法保持模型词汇大小不变，并在构建时作为兼容性属性保留大多数现有的标记到 ID 的映射。从特定语言的标记器直接转移标记并不能保证通过目标 BPE 合并图的可导性：插入的条目可能与目标的贪婪合并排名发生冲突。我们将这种失败形式化为合并顺序问题，并引入了 BPE 引导插入，该方法通过目标可达的分解构建每个转移的标记。我们的管道使用脚本感知的行选择来限制附带碎片，重建目标脚本的字节级先决条件，并应用引导插入以保持合并图的可达性。在乌克兰语对 Nemotron 和 GPT-OSS 的适配中，它将标记数量减少了 33.5% 和 36.6%，在英语和评估的四种语言的欧洲聚合中保持变化在 0.05% 以内，并在相同 ID 下保留了 78.5%/77.3% 的原始模型词汇行。约束匹配的全局和基于频率的移除实现了类似的乌克兰语压缩，但将英语/欧洲的标记数量增加了 0.7-2.2%；新的相同规模再训练稍微压缩了乌克兰语，但几乎没有保留相同 ID 的行，并将英语标记数量增加了 7.6-8.6%。重新分配使评估的三种语言的西里尔字母微聚合中的标记数量增加了 6.7%/10.1%。结构审计发现所有 28,134/45,398 个插入的 BPE 节点在普通排名顺序合并下可达，并且没有保留的相同 ID 模型词汇条目被新破坏。我们发布了所有标记器和代码。
Verification Without Sufficiency: Per-Chunk Filtering Fails on Multi-Hop RAG, and Decomposition Repairs It
没有充分性验证：逐块过滤在多跳检索增强生成中失败，而分解修复了这一问题
Kumar, Randhir
Abstract
Verification for retrieval-augmented generation usually scores each retrieved chunk and drops the ones that fail. We show this cannot work for multi-hop questions, and show what does. Per-chunk scoring assumes one chunk is a sufficient premise for the answer. Multi-hop questions are built so that none is, and the paragraph carrying the answer is the one the question does not name. Entailment scoring reaches 0.643, 0.523 and 0.560 AUC on HotpotQA, 2WikiMultihopQA and MuSiQue, against 0.951 on single-hop SQuAD. Seven controls rule out model capacity, premise length, hypothesis template, decision threshold, retriever, answer-matching criterion and prompt. End to end across three datasets, three generator sizes and two prompts, per-chunk gating is significantly worse than not filtering at all in every cell, and its penalty grows with generator capability. The repair is to condition verification on the decomposed sub-question rather than the original query. Using MuSiQue's gold decomposition, entailment on a later hop rises from 0.546, which is chance, to 0.840, a paired lift of +0.355 with a bootstrap interval of [0.331, 0.382]. An off-the-shelf Qwen2.5-7B decomposer, given the question and the top retrieved paragraph, reaches 0.637 and captures 31% of that ceiling; decomposing without retrieval reaches 0.533, below the original question. Iterative retrieval systems already produce such decompositions and discard them before verifying.
Language models (LMs) often acquire various biases during pre-training and may express them in interactions, potentially causing social harm. Existing methods often rely on counterfactual augmentation or representation projection. These strategies remain limited in practice due to their high computational costs and difficulty in scaling to larger models. Additionally, many of these strategies require manual data annotation, narrowing their scope to specific cultures and bias categories. To overcome these limitations, we propose HEIMAT, a HEurIstic-style autoMATic debiasing framework for LMs. HEIMAT consists of two main steps: bias disclosure and debiasing fine-tuning. In the first step, it uses simple templates to construct heuristic prompts, which are applied to reveal model biases and generate corresponding context prompts. In the second step, it fine-tunes the model by minimizing the Jensen-Shannon divergence of predictions on these context prompts to reduce bias. Extensive experiments show that HEIMAT effectively mitigates bias in different cultures while maintaining the model's natural language understanding (NLU) performance.
Large language models are increasingly viewed as a potential means of mitigating global health inequities, yet their outputs often reflect dominant high-resource medical traditions and provide limited coverage of traditional medical knowledge systems. Tibetan medicine, one of the world's four major traditional medical systems, has an independent and highly structured theoretical framework. When models lack grounded understanding of Tibetan medicine, they may fall back on dominant epistemic systems and distort the native knowledge structure during reasoning. However, quantitative tools for evaluating cultural bias in Tibetan medicine remain largely absent. To address this gap, we introduce TreeProbe, the first cultural-bias benchmark organized around the native Tree of Medicine framework in Tibetan medicine. It contains 4,719 expert-adjudicated items covering 467 diseases and 10 subtasks along the three roots. Experiments on representative LLMs show that current models remain limited in native Tibetan medical contexts and exhibit systematic external ontology drift. Further analysis reveals that models diverge in whether they drift toward biomedical or TCM reasoning, shaped by pretraining data composition and surface resemblance between TCM and Tibetan medicine. TreeProbe provides a diagnostic benchmark for developing medical AI systems that are both linguistically inclusive and epistemically fair. Code and data are available in an anonymous repository at https://anonymous.4open.science/r/TreeProbe/.
Select-And-Extract: A Lightweight Plugin for Retrieval-Augmented Generation
选择与提取：一种轻量级的检索增强生成插件
Tang, Chenming, Han, Jiawei
Abstract
Retrieval-augmented generation (RAG) for language model (LM) systems fundamentally has two failure modes: retrieval failure and reading failure. The former fails to recall the right pieces of information from the external corpus, and the latter fails to produce the correct answer although the right information is retrieved. Some methods perform structured indexing for retrieval failure, but may suffer from limited generalization of the fixed structures. Some methods perform query-time structuring for reading failure, but typically require a lot of LM calls and rely heavily on the LM's capability. To this end, we propose Select-ANd-Extract (SANE), a simple yet effective plugin for RAG. For the retrieval failure, we retrieve a wide set of candidates with a semantic retriever, and leverage the LM to select the top candidates based on their synopses, which yields better recall than the original retriever. For the reading failure, we perform blueprint-guided query-time evidence extraction, which allows the generator LM to use only compact and structured key information so that it can perform better reasoning. Empirical results confirm that SANE brings solid improvements, while only introducing modest extra overhead. As a lightweight plugin for RAG, SANE offers a simple alternative to heavier approaches, and suggests a high-performance RAG framework need not be overly complex.
AI agents operate in persistent environments where early state changes can influence decisions far into the future. Unlike conventional language-model interactions, agent behavior is mediated through a shared state that is repeatedly modified and reused across long-horizon workflows. Current safety benchmarks often fail to capture these cumulative risks because they focus on short, static tasks. To address these limitations, we introduce OpenART, an open-ended arena for scalable agent red teaming through environment evolution. OpenART provides over 10,000 validated stateful scenarios across 50 domains, drawing from a pool of more than 500,000 tools and skills. These tasks require a median of 97 tool calls and enable unified evaluation across 75 different agent-model configurations. To systematically explore these evolving attack surfaces, we propose the Evolutionary Markov Hypergraph Attack (EMHA). EMHA is a black-box policy that performs feedback-driven environment evolution by coordinating authorized state transitions without requiring parameter updates. Throughout the evaluation, task objectives remain fixed while only the environment state changes. Across all configurations, EMHA achieves a pooled Attack Success Rate (ASR) of 85.0%. Its advantage over instruction-only evolution increases from approximately 2% on simple environments to over 17% on the most complex ones, demonstrating that environment evolution increasingly exposes safety failures as task complexity grows. Furthermore, our analysis shows that the specific runtime implementation of an agent explains a significant portion of safety variation beyond the underlying model's capabilities. These results establish OpenART as a scalable foundation for studying agent safety in complex, evolving environments.
Schema linking is a critical component of Text-to-SQL systems, but existing approaches often trade off contextual modeling capacity, score-based controllability, and inference efficiency. We introduce AttnLink, an attention-based framework that converts LLMs' internal attention into continuous relevance scores for schema items. AttnLink extracts the attention from the generation-start position to candidate schema spans, enabling all candidates to be ranked in a single prefill pass without autoregressive decoding. We develop two variants: AttnLink-U, which directly probes pretrained attention without parameter updates, and AttnLink-S, which aligns the attention distribution with gold schema items through direct supervision. To improve coverage of multiple relevant schema items, AttnLink-S combines a set-mass objective with an adaptive probability-floor regularizer. The resulting scores support post-hoc precision-recall control through temperature scaling and cumulative-mass selection. Experiments on Spider, BIRD, and Spider2-SQLite show that AttnLink-S achieves mAP scores of 99.22%, 95.95%, and 83.29%, respectively, with millisecond-scale schema-linking latency. It also yields the best or tied-best execution accuracy for downstream SQL generation in seven of nine generator-dataset settings.
Multi hop question generation (MQG) aims to generate questions from multiple given documents and target answers, whereas question answering (QA) focuses on deriving answers from documents given specific questions. Although MQG and QA are inherently dual tasks, most existing MQG studies largely overlook this intrinsic duality. To address this limitation, we propose QQ, a novel framework that exploits the duality between Question and answer for multi hop Question generation. Specifically, QQ employs a unified architecture functioning simultaneously as both an MQG and a QA model to fully leverage their interdependence. Our framework is driven by two key mechanisms: (i) enforcing bidirectional alignment constraints to ensure strict mutual correspondence between the questions generated by the MQG model and the answers produced by the QA model; and (ii) applying contrastive learning to pull paired question answer representations closer while pushing unpaired ones apart, thereby reinforcing this correspondence. Extensive automatic and human evaluations on the HotpotQA and MuSiQue datasets demonstrate that the QQ framework significantly improves the quality of generated multi hop questions.
Observatorio Lazaro: A self-populating database of anglicism usage in the Spanish press
拉萨罗观察站：西班牙媒体中英语借用的自我填充数据库
Alvarez-Mellado, Elena
Abstract
This paper describes Observatorio L\'azaro, a language resource that monitors unassimilated lexical borrowings (predominantly English lexical borrowings or anglicisms) in the Spanish digital press. Since April 2020 the system has automatically processed the daily output of a collection of news outlets, detected borrowings with a neural sequence-labeling model, and made the results available through a public web interface and API. The result is a continuously updated diachronic database which, at the time of writing, records more than two million borrowings across 1.88 million articles and 993 million running tokens of text (2020-2026). The paper documents the resource: we describe the end-to-end pipeline (acquisition, detection, post-processing, storage and access), the data model and the terms of availability; we evaluate the resource through the detector's held-out performance (span-level F1=0.86 for the borrowing class), inter-annotator agreement on the training corpus (Cohen's kappa=0.91) and a manual precision audit of 1,000 spans from the deployed data; and we situate it with respect to Spanish borrowing lexicography, annotated borrowing corpora and neology-monitoring observatories. The data shows that unassimilated anglicisms are used in the Spanish press at a frequency of approximately two anglicisms per thousand tokens, and that this rate remains stable. Our statistical analysis over six years reveals that the anglicism vocabulary in Spanish behaves as an open and growing class, with 58.7% of its types attested only once (53.6% after correcting for detection precision), and that its density is highest in the fashion, technology and lifestyle sections and lowest in political and institutional news. The resource is intended to complement static borrowing dictionaries and one-off annotated corpora by providing a continuously updated record of borrowing in the Spanish press.
RAGOCR: Optical Compression of Retrieval-Augmented Text via Visual Representation
RAGOCR：通过视觉表示对检索增强文本进行光学压缩
Yu, Jiayang, Zhong, Jialun, Zou, Lei
Abstract
Retrieval-Augmented Generation (RAG) has become essential for knowledge-intensive question answering, yet scaling RAG pipelines remains challenging due to the prohibitive computational cost of processing lengthy retrieved contexts. Existing compression approaches face a fundamental trade-off: hard compression methods operate online in a query-aware fashion but achieve only modest compression rates and typically require fine-tuning the generative model, while soft compression methods attain higher ratios but rely on costly offline encoding that is entirely agnostic to the input query. To bridge this gap, we introduce RAGOCR, a novel framework that compresses retrieved documents into compact visual representations conditioned on the input query. To further balance compression rate and information fidelity, we introduce a query-aware dynamic resolution mechanism that adaptively allocates visual granularity based on each document's estimated relevance and complexity: highly relevant passages are rendered at higher resolution to preserve fine-grained details, while peripheral documents are aggressively compressed at lower resolution. Experiments on five QA benchmarks using the MedOmniKB retrieval corpus demonstrate that RAGOCR surpasses naive RAG by over 15\% in accuracy while requiring only one-eighth the number of input tokens, and consistently outperforms both hard and soft compression baselines across varying retrieval depths.
Reinforcement learning with verifiable rewards (RLVR) has become a standard paradigm for post-training large language models (LLMs). While Group Relative Policy Optimization (GRPO) is widely adopted, it suffers from sparse reward signals and loses gradients entirely when all responses within a group receive identical rewards. On-policy distillation (OPD) offers a natural remedy by providing dense, token-level supervision from a teacher model. However, naively combining GRPO with OPD leads to degraded performance, due to three underlying causes: not all samples benefit from distillation; fitting too quickly to the teacher undermines the exploratory capacity of RL; and OPD's advantages are asymmetric, suppressing most tokens. To address these challenges, we propose RSTG (Recovering Learning Signals via Adaptive Teacher Guidance), which applies distillation selectively and precisely where it matters most. At the sample level, OPD is restricted to negative zero-variance prompts with each sample weighted by the teacher's confidence score. At the token level, distillation targets only tokens with high student entropy or large teacher-student divergence. We further augment training with SFT on correct trajectories generated by the teacher model, injecting positive gradient signals where RL yields none. Experiments demonstrate that RSTG substantially outperforms naive GRPO+OPD by +4.02% on math and +3.05% on code.
LLMs generate tool calls token by token, even though the function choice and argument values can often be predicted in parallel from the request and tool schema. ToolSpec reduces this cost by drafting schema tokens and retrieving earlier calls, but cannot propose request-specific values absent from either source. We present OoO-Spec, which computes these missing semantics out of order. At request arrival, a Qwen3-0.6B sidecar predicts the function choice and all schema-defined argument slots in one parallel request-level wave while the target begins ToolSpec decoding. The runtime joins the slot values, renders the resulting call as text, and exposes it to subsequent candidate-construction rounds. The target polls without blocking, re-tokenizes a ready hint with its own tokenizer, and remains the sole verifier and commit authority. The sidecar is trained once with LoRA on Qwen2.5-32B teacher traces and used unchanged across Qwen2.5, Qwen3, and Llama targets, without target-specific drafter training. Across seven fully ranked targets and three benchmarks under greedy batch-one decoding, OoO-Spec is fastest among all evaluated methods in all 21 target-benchmark cells, reaching 2.46x-5.34x over autoregressive decoding with an unweighted mean of 3.89x, versus 2.95x for ToolSpec. It also outperforms every evaluated released learned drafter in each comparable cell. Across Qwen3-4B, 8B, 14B, and 32B targets, the same sidecar improves on ToolSpec by 34.1% on average. Its compact semantic payload averages 85 bytes per request excluding protocol metadata, supporting effective split-GPU overlap.
Exemplars in Disguise: Pure Exemplar Models Mimic Abstraction-First Learning
伪装中的典范：纯典范模型模拟抽象优先学习
Houghton, Zachary Nicholas, Kapatsinski, Vsevolod
Abstract
Whether idiosyncratic, item-specific knowledge is learned before abstract class-level generalizations, or vice versa, is a central question in language learning, with exemplar and abstraction-based theories making opposite predictions. Recent methods have claimed to show that, at least for large language models, abstract knowledge is learned first. We show that these methods fall short: pure memorizer models with no abstract representations can appear, by the same criteria, to learn either item-specific or class-level knowledge first, depending on their sensitivity to individual observations, with the transition point governed by the distributional properties of the input. We further argue that the distinction between item-specific and abstract knowledge may be ill-defined for distributed representations, as a word's class-level properties may not be separable from its item-specific properties.
Pruned BPE: Post-training Visibility Pruning and Token Reallocation for Byte Pair Encoding
修剪的字节对编码（Pruned BPE）：后训练可见性修剪与标记重新分配
Shao, Kenny
Abstract
Byte Pair Encoding (BPE) is widely used for subword tokenization, but standard BPE exposes every learned merge token to the downstream model, including tokens that mainly serve as intermediate construction units and rarely appear in the final encoded corpus. This paper proposes Pruned BPE, a post-training visibility-pruning and token-reallocation method that separates merge construction from model-visible vocabulary selection. After standard BPE training, tokens are evaluated by final exposure. Low-exposure tokens are retained as internal-only merge nodes, while their visible vocabulary slots are reassigned to better-exposed candidates learned through resumed training. During encoding, internal-only tokens are recursively expanded into visible descendants while the original BPE merge order is preserved. Experiments on two non-overlapping English- and Chinese-dominated corpora and their combination show that Pruned BPE consistently reduces encoded length relative to Standard BPE at the same training corpus, evaluation corpus, and model-visible vocabulary size. At a 40\% exposure threshold, the reduction is approximately 0.27\%--0.36\% on same-corpus evaluations. In a vocabulary-only evaluation using a shared exact minimum-token dynamic-programming encoder, Pruned BPE retains an advantage of approximately 0.23\%--0.31\%, indicating that the improvement arises from a more efficient visible vocabulary. These gains represent a meaningful fraction of the approximately 1.5\%--3.8\% marginal reduction that would otherwise require adding another 2K Standard BPE tokens. Qualitative analysis shows that internal-only tokens include reusable English fragments, Chinese components, partial UTF-8 byte sequences, and structured-text fragments. The results indicate that post-training visibility pruning can improve BPE vocabulary efficiency without increasing the vocabulary exposed to the language model.
Practical Online KV Cache Compaction for LLM Agents: An Empirical Study
针对LLM代理的实用在线KV缓存压缩：一项实证研究
Liu, Yujian, Ji, Jiabao, An, Li, Jain, Rohit, Polatkan, Gungor, Zhu, Siyu, Chang, Shiyu
Abstract
LLM agents accumulate long trajectories of reasoning steps, tool calls, and environment feedback, making the KV cache a major inference bottleneck. KV cache compaction can reduce this cost, but most prior methods assume a static context where future queries are known or can be approximated offline. Agents instead require online compaction: new information must be compressed before future relevance is known, using proxy queries cheap enough for the inference path. We study online compaction across token eviction (TE) and attention matching (AM), adapting both to compact agent turns and comparing cheap proxy sources such as boundary, repeat-prefill, and delayed future-generation queries. Experiments on BrowseComp-Plus and WideSearch show that immediate compaction often hurts performance, whereas delaying compaction to use the agent's future queries recovers much of the gap. Moreover, TE is often more robust than AM under imperfect proxies. Across models at different scales, TE preserves most of the accuracy while reducing KV cache by 80%, and can improve throughput over the no compaction baseline. These results position proxy-query selection as a core design choice for practical online KV compaction.
Can large language models generate not just correct, but fast hardware? This paper investigates the question in financial FPGA design, where 5-10 nanoseconds of latency determines competitive advantage and designs iterate continuously as protocols, strategies, and regulations evolve. FinHardBench, a benchmark of 33 financial computing tasks, is presented together with three experiments that mirror the real-world FPGA iteration cycle: generating new modules from specifications, tuning system-level configurations across a 6-stage trading pipeline, and adapting existing modules to specification changes. Evaluation of six LLMs on 1530+ experiment rounds yields three findings: (1) models achieve 19-61% functional correctness with timing degradation up to 13.7$\times$ on specific tasks; (2) in system-level design space exploration, top LLMs converge to the optimal configuration with higher reliability than random search, simulated annealing, and Bayesian optimization baselines (5/5 seeds vs. 0-4/5 at the same 24-round budget); (3) strategy-level specification changes remain unsolved for most models. Across the six models, generation and DSE rankings overlap moderately: the strongest code generator is not the fastest architecture optimizer, and the weakest code generator (MiniMax M2.7) still reaches the system optimum on 4 of 5 seeds. On the tasks in FinHardBench, difficulty tracks training data pattern availability more closely than abstraction level. FinHardBench is released as an open-source benchmark.
The integration of artificial intelligence into medical question-answering systems has advanced rapidly; however, research remains predominantly focused on English, leaving low resource languages like Persian significantly underserved. To address this gap, this paper introduces Gaokerena, a novel family of compact Persian medical language models optimized for deployment on consumer grade hardware. As a foundational step toward localized digital healthcare, we first present Gaokerena-V, developed by training a baseline model on a newly curated 90-million-token Persian medical corpus and 20,000 expert-vetted physician Q&A pairs, which improved performance on a translated medical MMLU benchmark from 46.28% to 49.31%. Second, recognizing the critical demands of clinical reasoning, we developed Gaokerena-R by integrating a Chain-of-Thought approach with two novel Reinforcement Learning with AI Feedback (RLAIF) frameworks to optimize preference-based reasoning. Despite utilizing the same baseline architecture and a smaller dataset than Gaokerena-V, Gaokerena-R achieved a superior benchmark score of 52.98%. Furthermore, both models are equipped with custom-developed uncertainty heads that predict the model's confidence in its responses based solely on internal hidden states. While these results demonstrate significant progress in Persian medical language modeling and proactive safety estimation, current performance levels remain insufficient for direct clinical application, highlighting the necessity for further research into robust knowledge acquisition and rigorous safety verification prior to real world deployment.
Chinese Translation
人工智能在医学问答系统中的集成迅速发展；然而，研究仍主要集中在英语上，导致波斯语等低资源语言受到显著忽视。为了解决这一问题，本文介绍了Gaokerena，一个新颖的紧凑型波斯医学语言模型系列，旨在优化在消费级硬件上的部署。作为本地化数字医疗的基础步骤，我们首先提出了Gaokerena-V，该模型通过在新近整理的9000万标记的波斯医学语料库和20000对经过专家审核的医生问答对上训练基线模型而开发，性能在翻译的医学MMLU基准测试中从46.28%提升至49.31%。其次，考虑到临床推理的关键需求，我们通过将Chain-of-Thought方法与两个新颖的基于强化学习的AI反馈（Reinforcement Learning with AI Feedback, RLAIF）框架结合，开发了Gaokerena-R，以优化基于偏好的推理。尽管使用了与Gaokerena-V相同的基线架构且数据集规模更小，Gaokerena-R仍然达到了更高的基准得分52.98%。此外，这两个模型都配备了定制开发的不确定性头部，能够仅基于内部隐藏状态预测模型对其响应的信心。尽管这些结果展示了波斯医学语言建模和主动安全估计的显著进展，但当前的性能水平仍不足以直接应用于临床，突显了在实际部署之前对稳健知识获取和严格安全验证进行进一步研究的必要性。
Mind the Gap: Zero-Query Jailbreaks via Filter-Generator Discrepancy in Text-to-Image Systems
注意差距：通过文本到图像系统中的过滤器-生成器差异实现零查询越狱
Li, Wanguang, Wang, Zhaoxin, Wang, Handing
Abstract
Text-to-image (T2I) systems typically have prompt-level safety filters before the generator to block unsafe requests, yet such systems remain vulnerable to malicious jailbreak prompts. Transfer-based attacks construct adversarial prompts offline without querying the target, but they tend to overfit to a single surrogate. Moreover, they explore a large search space in which semantic or perceptual similarity alone cannot guarantee both filter evasion and preservation of the unsafe generation intent, wasting effort on low-potential candidates. We observe that the filter and the generator process the same prompt under different objectives and representations, and term this gap the Filter-Generator Discrepancy (FGD), which allows a perturbation to reduce a prompt's perceived risk to the filter while preserving the visual concept needed by the generator. Building on FGD, we propose a zero-query jailbreak framework that screens perturbations into a high-potential candidate set via observable discrepancy rules at the tokenization and semantic stages, and then performs a surrogate-ensemble evolutionary search that requires no access to the target. Experiments on six black-box pipelines and a commercial online service show that our method consistently outperforms representative baselines, raising the average attack success rate to 29.2\% (MHSC) and 33.3\% (Q16) across the six pipelines and improving over the strongest baseline by about 8 and 12 percentage points, respectively.
Unsupervised Multidomain Approaches to Named Entity Recognition with Small Datasets
小数据集下无监督多领域命名实体识别方法研究
Fianyi, Israel, Montgomery, James, Yeom, Soonja
Abstract
This paper explores the challenges and the methodologies associated with learning quality representations in scenarios with unlabelled small or limited datasets for downstream information extraction task (Multidomain Named Entity Recognition (NER). The study adopts a Transfer Learning on small datasets. Traditional NER systems often rely on large, labelled data, which is impractical for many domains. This study, therefore, applies an unsupervised pre-training approach to precondition and identify entities without annotated datasets, then applies transfer learning models to different simulated limited datasets for a named entity recognition task. Entity Recognition (NER) is essential in natural language processing (NLP), it identifies and classifies related entities within the text. This study addresses the complexities of domain variability, data sparsity, and overfitting and investigates innovative approaches such as data augmentation, few-shot learning, and domain adversarial training. Integrating these techniques promises to enhance the performance and generalizability of NER systems across diverse and resource-constrained domains, paving the way for more efficient and adaptable NLP applications.
Uncommon and off-guideline cases are difficult for clinical decision support, because physicians must make a series of management decisions under diagnostic uncertainty and rarely see the full case at once. Most large language model (LLM) benchmarks for medicine score only the final diagnosis, yet much of clinical care turns on the next appropriate action: the next test to order, the imaging study to obtain, the specialist to involve, or the differential to pursue. We introduce MedUPSQA, a dataset of 21,874 mid-stream clinical decision points built from 5,535 real case reports, and MedUPS, an alignment framework that supervises models on these intermediate decisions as they unfold along a patient's trajectory. We segment free-text case presentations into chronologically ordered, accumulating clinical chunks and align models to predict the next step with reinforcement learning (GRPO), using an external LLM-as-a-Judge reward. This objective mirrors how clinicians actually meet patients, reasoning forward from accumulating evidence toward the next decision, rather than committing to a final label. Across three backbones, mid-stream alignment raises next-step accuracy from 55.2 to 66.7 for Qwen3.6-27B, from 47.2 to 57.8 for Qwen3.5-9B, and from 37.8 to 44.4 for HuatuoGPT-3-8B, with 95% CI. In several model scales we test the objective improves accuracy more than scale, with smaller models surpassing larger, frontier models we evaluate. We further train supervised fine-tuning (SFT) baselines on the mid-stream task, SFT improves all backbones above base, indicating the target framwork carries signal independently of the optimizer. We release the dataset, code, and aligned checkpoints.
Cloud-ScPO: Hidden-State Geometry for Semi-Supervised Preference Optimization in LLM Reasoning
Cloud-ScPO：大语言模型推理中的半监督偏好优化隐状态几何
Liu, Yuzhou, Hu, Xiyang
Abstract
Preference optimization improves mathematical reasoning in large language models (LLMs), but reliable chosen-rejected pairs usually require verified answers, human annotations, or external reward models. We investigate whether preference supervision can instead be derived from the model's internal representation geometry in a semi-supervised setting. Our analysis shows that reasoning trajectories generated across different mathematical problems form structured global point clouds in which correct and incorrect trajectories exhibit different geometric organization. Based on this observation, we propose Cloud--ScPO, a topology-guided preference-mining framework that uses a small labeled set to construct multiple correct and incorrect reference Clouds. Each trajectory is represented by a mean-pooled hidden state and scored against connectivity-induced components using a component-level soft $k$-nearest-neighbor measure averaged across reference banks. We combine this cross-problem Cloud signal with prompt-level self-consistency: self-consistency determines the answer-level preference direction, while Cloud scoring selects concrete trajectories and filters pairs by their score margin. Experiments on GSM8K and MATH-Numeric across four model settings show that Cloud--ScPO consistently improves over ScPO, with gains of up to 4.49\% on GSM8K and 4.19\% on MATH-Numeric. Pair-level analyses further show that Cloud--ScPO maintains comparable correctness reliability while more effectively separating informative chosen trajectories from incomplete, repetitive, or otherwise low-quality rejected responses.
A language model that abandons a correct medical answer under user pushback is more dangerous than one that was simply wrong, because it lends the credibility of a correct answer to the user's misinformation. Such model behavior, described as medical sycophancy, is usually reported as a single rate per model, but we find it is a property of the conversation, not the model. We study medical sycophancy in language models with a fully crossed factorial design over four conversational factors, user role, the evidence behind a false claim, whether the challenge precedes or follows the model's answer, and whether the correct answer is grounded in the prompt, across five open-weight models and 500 MedQuAD questions (1.2M trials). The factors interact sharply: fabricated sources raise sycophancy 2.0x when they accompany the question but halve it once the model has answered, so the same evidence helps or hurts depending only on timing. Sycophancy varies far more across questions than across models (67x vs. 3x), so a single rate reflects the conversation and the questions sampled as much as the model. Chain-of-thought traces explain why. Models that re-examine their own prior answer concede, while those that reason about the medical facts hold, and only a model that has already answered can spend a round auditing the fabricated source.
We introduce Opt.Gear, a foundation model designed for efficient on-device deployment, real-tim inference, and strong task capability. It includes a dense model (1M, 270M, and 1B) with a context length of 64K. We designed a new hybrid architecture that combines a convolutional key-value gated mixer with local-global attention to reduce the KV-cache memory that tends to increase exponentially with long context. This architecture delivers up to X4.9 faster prefill and decoding speeds on the NPUs compared to models of a similar scale models. From a 2T tokens candidate corpus, Opt.Gear is trained on a curated 0.5T tokens subset without knowledge distillation. This is the most data-efficient of the existing foundation models. All models are released with open weights and deployment binaries for ONNX, Qualcomm NPU, and Apple ANE making Opt.Gear a practical base for edge applications that need fast, memory-efficient inference and strong task capabilities. Furthermore, to expand the ecosystem of on-device generative language models, we are introducing the Opt.Gear-1M that can be deployed on Micro-Controller Units (MCUs), a Tiny Language Model (TLM). Opt.Gear-1M is the first generative language model to achieve 20 TPS with W4A32 quantization on the ARM Cortex-M7 CPU of the STM32H747I-DISCO.
DeBERTa-Sentinel: Toward Transparent and Trustworthy Detection of AI-Generated Text
DeBERTa-Sentinel：迈向透明和可信的人工智能生成文本检测
Rehman, Muhammad Yousaf, Islam, Muhammad
Abstract
The rapid spread of large language models (LLMs) across the web raises concerns about misinformation, academic integrity, automated content manipulation, and risks to vulnerable online communities. Existing transformer-based detectors, such as GPT-Sentinel, show promise but struggle to generalize to diverse model outputs and paraphrasing attacks, limiting their role in building trustworthy web ecosystems. This work introduces DeBERTa-Sentinel, a responsible AI-generated text detection framework leveraging DeBERTa-v3's disentangled attention to capture subtle structural irregularities in synthetic content. A central design principle is transparency: unlike black-box commercial detectors, DeBERTa-Sentinel exposes token-level explanations of its decisions, enabling affected stakeholders journalists, educators, and platform trust and safety teams to audit, challenge, and contextualize detection outcomes. Using the GLC-AIText dataset of 28,057 human and LLM-generated samples (GPT, LLaMA, and Claude) with a 60-20-20 split, DeBERTa-Sentinel achieves 98.21\% validation accuracy and surpasses the RoBERTa-Sentinel baseline from NeurIPS 2025, achieving 97.53\% test accuracy, 95.89\% precision, 99.33\% recall, and 99.53\% ROC-AUC, and maintaining a 0.665\% false negative rate. The model's interpretability reveals linguistic markers such as academic phrasing and formal transitions associated with synthetic text, directly supporting stakeholder needs for verifiable, auditable content-authenticity decisions. By advancing responsible detection methods that reduce bias and enhance explainability, DeBERTa-Sentinel promotes trustworthy, ethical, and human-centric AI systems. Code and data are available at https://github.com/Galileo-Galili/HUMAN-VS-AI-TEXT-DETECTION.
We propose ScaleQ-1.58, a scalable ternary post-training quantization (PTQ) framework for reasoning LLMs. Its core insight stems from an empirical finding: although modern LLMs are typically trained to exhibit chain-of-thought reasoning capabilities, in the PTQ regime, even the latest CAT-Q method based on learning-based differentiable ternarization still leads to performance collapse on challenging mathematics and coding tasks when using conventional calibration schemes that ignore the model's reasoning process. Driven by this finding, we introduce a simple calibration approach, Attend to Your Own Thoughts (AYOT), where reasoning traces and final answers generated by the pre-trained high-precision target LLM on a proper set of calibration samples are used as the context input during the ternarization process, along with the corresponding questions. ScaleQ-1.58 is formed by simply integrating AYOT with CAT-Q, which demonstrates several scaling properties: (1) with only 4M calibration tokens, Qwen3-1.7B ternarized by ScaleQ-1.58 reaches over 90.52% of the performance of the prior best BitNet b1.58 2B4T averaged over 4 mathematics and coding tasks, and our ternary Qwen3-4B shows an absolute gain of 8.97%, while requiring 1,000,000x fewer calibration tokens for quantization; (2) ScaleQ-1.58 generalizes well to both dense and MoE architectures, with performance improving as model scale increases (up to 235B parameters); (3) ScaleQ-1.58 demonstrates strong generalization across tasks of varying difficulty levels, including mathematics, coding and scientific logic reasoning, as well as commonsense reasoning and basic language generation; (4) its performance continues to improve as the number of calibration tokens increases. Notably, AYOT also exhibits strong generalization ability across other quantization bit-widths. Code will be available at https://github.com/IntelChina-AI/BitTern.
Chinese Translation
我们提出了ScaleQ-1.58，这是一个可扩展的三元后训练量化（PTQ）框架，旨在用于推理大型语言模型（LLMs）。其核心见解源于一个经验发现：尽管现代LLMs通常被训练以展现链式思维推理能力，但在PTQ环境下，即使是基于学习的可微分三元化的最新CAT-Q方法，在使用忽略模型推理过程的传统校准方案时，仍会导致在具有挑战性的数学和编码任务上的性能崩溃。基于这一发现，我们引入了一种简单的校准方法，称为关注你自己的思维（Attend to Your Own Thoughts, AYOT），在三元化过程中，使用预训练的高精度目标LLM在适当的校准样本集上生成的推理轨迹和最终答案作为上下文输入，同时结合相应的问题。ScaleQ-1.58通过简单地将AYOT与CAT-Q结合而形成，展示了几个扩展特性：（1）仅用4M校准标记，ScaleQ-1.58量化的Qwen3-1.7B在4个数学和编码任务上达到了超过90.52%的前最佳BitNet b1.58 2B4T的性能，而我们的三元Qwen3-4B则显示出8.97%的绝对增益，同时在量化时需要的校准标记少达1,000,000倍；（2）ScaleQ-1.58在稠密和混合专家（MoE）架构上均表现良好，随着模型规模的增加（最多达到235B参数），性能不断提升；（3）ScaleQ-1.58在不同难度级别的任务上展现出强大的泛化能力，包括数学、编码和科学逻辑推理，以及常识推理和基础语言生成；（4）随着校准标记数量的增加，其性能持续改善。值得注意的是，AYOT在其他量化位宽上也表现出强大的泛化能力。代码将可在https://github.com/IntelChina-AI/BitTern获取。
Morphology Aware Reversible Semantic Tokenization and Hierarchical Word Composition for Tamil Language Models
考虑形态的可逆语义分词与层次化词汇组合在泰米尔语言模型中的应用
Murugan, Anand
Abstract
Statistical subword tokenizers can process arbitrary text, but their units need not align with lexical or grammatical structure. This is especially important for Tamil, where a written word may encode stem changes, case, number, tense, agreement, voice, clitics, and linked verbs. We present a Tamil morphology system extending the open-source ThamizhiMorph analyzer and generator, together with a byte-exact semantic tokenizer and a learned hierarchical word composer. Twelve finite-state transducers analyze words into lemmas and grammatical features, while character and byte fallbacks preserve exact reconstruction. We compare a flat morphology tokenizer, a signal-preserving word composer, and tokenizers based on Sarvam-1, AI4Bharat IndicBERTv2, and BrahmicTokenizer-131K. All systems use the same 69,591 Tamil-English training pairs, 18.97-million-parameter encoder-decoder, 40,000 updates, target tokenizer, optimizer, positional method, and generation settings. On a protected 3,539-row IN22 and FLORES+ evaluation, morphology-flat achieves the best pooled scores: 10.63 BLEU, 35.26 chrF++, and 0.6276 COMETKiwi. Relative to AI4Bharat, the strongest external-tokenizer baseline, these are improvements of 7.2%, 3.2%, and 2.6%. The word composer scores 10.30, 34.88, and 0.6241, improving on AI4Bharat by 3.8%, 2.1%, and 2.0%. The composer reduces mean global source states from 71.48 to 29.08, a 59.3% reduction, and is estimated to require 9-21% fewer inference FLOPs depending on decoder caching. Its remaining quality gap is concentrated in longer FLORES+ sentences. These results show that explicit Tamil morphology improves translation under a fixed small-model budget, while hierarchical composition substantially reduces sequence length and estimated inference cost.
PlainMedScale: A Corpus of Multi-Level Simplified Medical Texts in German and English
PlainMedScale：德语和英语多层次简化医学文本语料库
Brocai, Bruno, Papagno, Ilaria, Ohta, Mayumi
Abstract
We introduce PlainMedScale, a topic-aligned medical corpus spanning four levels of comprehensibility in German and English, drawn from MSD (professional and consumer), Gesund.Bund, Apotheken Umschau Einfache Sprache, and the NHS. The four tiers correspond to distinct communicative functions --- reference, explanation, decision support, and access --- and move beyond the binary expert--lay contrast of prior corpora. In two pilot studies enabled by the alignments, we show that many readability metrics established on two registers fail to generalize across the full gradient, and that a SOTA open-weight LLM prompted for Plain Language still partially preserves the difficulty of its input. Code (https://github.com/GS-Uni-Heidelberg/PlainMedScale) and data (https://doi.org/10.5281/zenodo.21728290) are made available.
Does Machine "know" interpersonal pragmatics? Evidence from MARBERT's learning of emoji pragmatics in Arabic digital discourse
机器是否“理解”人际语用学？来自MARBERT在阿拉伯数字话语中学习表情符号语用学的证据
Shormani, Mohammed Q.
Abstract
This study examines Transformer-based models' ability to learn emoji pragmatics in Arabic digital discourse (ADD), providing evidence from MARBERT's behavior with interpersonal pragmatic functions (IPFs). A corpus of 8,504 unique emoji-posts collected from Facebook via Python was used in the study. These posts were manually annotated, developed, and labeled for five IPFs: Politeness, Respect, Solidarity, Empathy, and Encouragement. A mixed-method approach was employed comprising statistical methods and interpretative analyses involving speech act theory, politeness theory, and rapport management theory. MARBERT was fine-tuned to model these context-dependent pragmatic functions. Findings demonstrate MARBERT's ability to learn these IPFs, achieving strong performance on unseen data, with an accuracy of 93%, a micro F1-score of 0.61, and a macro F1-score of 0.56, demonstrating its effectiveness in capturing interpersonal functions beyond conventional sentiment analysis. Function-level evaluation showed that Politeness and Respect were identified more accurately than Solidarity, reflecting differences in the explicitness and contextual dependence of IPFs. The study concludes that Transformer-based models learn patterns of face management and relational communication but remain challenged by highly implicit social meanings. It contributes a novel computational approach to modeling emoji pragmatics and advances the integration of interpersonal pragmatics with NLP for digital communication research.
When Words Divide: Diachronic Ideological Polarization in Political Discourse on Social Media
当言语分裂：社交媒体政治话语中的历时意识形态极化
Yitzchak, Roy, Lavie, Noa, Rabinovich, Ella
Abstract
Political polarization has become a defining feature of online discourse, yet its long-term evolution remains poorly understood. We present a longitudinal analysis of ideological polarization in Reddit discussions by measuring semantic differences in the language used by opposing political communities. We construct temporally aligned community-specific word embeddings and quantify ideological polarization as the semantic divergence of political concepts over time. Our analysis shows that ideological polarization has increased substantially during the study period, both at the concept- and topic-level. Unlike prior computational work, which has largely focused on cross-sectional analyses or affective dimensions of polarization at a single point at time, our approach captures the evolution of ideological differences in semantic framing. The proposed framework provides a scalable method for studying the temporal dynamics of ideological polarization in large-scale social media discourse.
Conversational investment advisors influence not only what users know, but also how they make subsequent decisions as market conditions evolve. Existing evaluations primarily assess response quality or observed outcomes, leaving the long-horizon pathway from advisor language to investor behavior difficult to audit. We introduce ShiJianBench, an offline framework for evaluating conversational investment advisors through matched investor trajectories under fixed historical market feedback. At its core is a multi-agent investor simulator with explicit evolving state variables, motive-driven deliberation, long-term memory, and dialogue-grounded updates. The simulator is calibrated against aggregate behavioral patterns from 7,199 real users, and advisor policies are evaluated using separate investor-side, service-side, and content-side metrics under a hard compliance gate. Experiments on Chinese fund-market traces from 2021 to 2026 identify a stable leading group of LLM advisors that combines substantially stronger personalized content with competitive investor-side trajectory outcomes. These results reveal a systematic distinction between producing a high-quality response and delivering an effective long-horizon intervention, motivating trajectory-aware evaluation of conversational advisors.
Evaluating VLMs on Multimodal Aristotelian Persuasion Tasks
在多模态亚里士多德说服任务上评估视觉语言模型
Islam, Khondoker Ittehadul
Abstract
Vision Language Models (VLMs) have demonstrated exceptional performance across various tasks. However, they have not yet been thoroughly evaluated on more complex tasks. The Persuasion Model, conceived by Aristotle, resembles a triangle shape, which highlights its inherent challenges related to personal biases. To assess the progress of VLMs on these complex tasks, we use the ImageArg datasets, focusing on the Logos, Ethos, and Pathos detection tasks. Our findings indicate that models from the Qwen family achieve improved F1 scores, with Qwen3 performing exceptionally well on the Logos and Pathos tasks, while Qwen2 exhibits competitive performance on the more complex Ethos detection task. We release the code to foster research in this direction.
Embedded Language Flows (ELF) rely primarily on full non-causal attention for iterative denoising, repeatedly incurring quadratic sequence-mixing cost at each sampling step. Gated Delta Networks (GDNs) provide an efficient recurrent alternative, but their standard causal formulation cannot directly capture the bidirectional context required by ELF. We introduce DeltaFlow, a noise-adaptive bidirectional GDN backbone for continuous language denoising. We study two variants: DeltaFlow-A, which alternates scan directions across layers, and DeltaFlow-P, which performs parallel forward and backward scans within each layer. We further introduce noise-adaptive memory control and scheduled Temporal State Consistency (TSC) to stabilize hidden representations across nearby noise levels. On OpenWebText, using a 32-step stochastic differential equation sampler, DeltaFlow-P reduces generated perplexity from 24.218 for the full-attention ELF baseline to 21.228 while maintaining comparable unigram entropy, with 36B training-token exposure compared with 45B for the baseline. In a denoiser-only benchmark, DeltaFlow-P achieves a 2.72x throughput speedup over the full-attention baseline at a sequence length of 16k. These results show that DeltaFlow is a promising alternative to dense attention for efficient continuous language denoising.
Chinese Translation
嵌入式语言流（Embedded Language Flows, ELF）主要依赖全非因果注意力进行迭代去噪，每次采样步骤都重复产生二次序列混合成本。门控Delta网络（Gated Delta Networks, GDNs）提供了一种高效的递归替代方案，但其标准因果形式无法直接捕捉ELF所需的双向上下文。我们提出了DeltaFlow，一种用于连续语言去噪的噪声自适应双向GDN骨干网络。我们研究了两种变体：DeltaFlow-A，它在各层之间交替扫描方向，以及DeltaFlow-P，它在每层内执行并行的前向和后向扫描。我们进一步引入了噪声自适应记忆控制和调度的时间状态一致性（Temporal State Consistency, TSC），以稳定相邻噪声水平下的隐藏表示。在OpenWebText数据集上，使用32步随机微分方程采样器，DeltaFlow-P将生成的困惑度从全注意力ELF基线的24.218降低到21.228，同时保持可比的单元熵，训练标记曝光量为36B，相较于基线的45B。在仅去噪器的基准测试中，DeltaFlow-P在序列长度为16k时实现了比全注意力基线高出2.72倍的吞吐量加速。这些结果表明，DeltaFlow是高效连续语言去噪的有前景的替代方案。
RestoreKV: Recovering Full-Cache Behavior Under Aggressive Query-Agnostic KV Cache Eviction
RestoreKV：在激进的查询无关 KV 缓存驱逐下恢复全缓存行为
Baek, Changwoo, Shin, Seungjun, Kong, Kyeongbo
Abstract
Query-agnostic KV cache eviction compresses a context once and reuses the resulting cache for arbitrary future queries, but performance can collapse under tight budgets. Existing methods primarily improve which original KV pairs are retained. We introduce RestoreKV, which complements this selection-based formulation with learned restoration under the same total KV budget. Our key insight is that, although the information lost through eviction is context-specific, the mechanism for generating its compact complement can be shared across contexts. After context prefill, a few restore tokens attend to the full KV cache in a single LoRA-adapted pass, generating a compact, context-conditioned restore cache. The base importance scorer and eviction rule remain unchanged, and the adapters are disabled for all subsequent queries and decoding. RestoreKV is trained through parameter-efficient self-distillation from the frozen full-cache model, optimizing only $0.4\%$ of the parameters and requiring no task-specific tuning. Across four backbones and four long-context benchmarks, RestoreKV substantially reduces compression-induced degradation. On Qwen3-4B, it improves 59 of 60 paired, budget-matched settings across five base eviction methods; at a $5\%$ budget, it raises KVzip from $38.2$ to $73.2$ on RULER-4K. Applied to KVzip+, RestoreKV reaches $86.4$ RULER accuracy at $16\times$ compression on the KVPress Benchmark, while adding less than $0.5\%$ one-time cache-construction overhead in a 32K-context evaluation. Our project page is available at https://paper.pnu-cvsp.com/RestoreKV/
Hierarchical Graph Retrieval-Augmented Generation (GraphRAG) organizes corpus knowledge at multiple levels of granularity, yet fixed context construction may fail to translate these multi-resolution representations into a context suited to the current query. We identify this mismatch as the representation--inference gap. We propose Agentic Context Engineering for Hierarchical GraphRAG (ACE-GraphRAG), an inference-time context policy layer that supplements and adapts the initial context for generation. ACE-GraphRAG formulates context construction as a policy over gap-aware refinement, retrieval branches, and task-conditioned adaptation. Parallel Differential Retrieval acquires supplementary evidence from depth-oriented factual and breadth-oriented semantic branches. These evidence increments are consolidated with the initial context while preserving provenance and abstraction levels. Full-ACE applies the full policy uniformly within each task family, whereas Adaptive-ACE selects task- and topology-specific policies for individual queries. We evaluate ACE-GraphRAG on HotpotQA, 2WikiMultiHopQA, and four UltraDomain subsets across multi-hop QA and query-focused summarization. Full-ACE outperforms the evaluated RAG and GraphRAG baselines across both task families, while Adaptive-ACE further improves multi-hop QA and is preferred over Full-ACE on all four UltraDomain subsets. Ablation and topology analyses support treating context construction as a query- and task-dependent inference policy rather than a fixed procedure.
We present ArabicDialectSafety, a human-curated Arabic safety dataset of 25,071 prompts covering six Arabic varieties: Modern Standard Arabic, Syrian, Egyptian, Algerian, Palestinian, and Moroccan. The dataset is annotated with dialect labels and seven fine-grained harm categories. We introduce a dual-task evaluation framework for binary safe/unsafe detection and granular harm classification across dialects. Benchmarking seven supervised and generative models, we find that fine-tuned MARBERTv2 achieves the strongest performance, with Macro-F1 scores of 0.95 for binary classification and 0.90 for granular classification, substantially outperforming prompted frontier LLMs, including Arabic-specialized models. Our analyses show that dialect conditioning is most effective when integrated at the representation level, while significant performance gaps remain for low-resource Maghrebi dialects. We further evaluate seven frontier LLMs as response generators on harmful dialectal Arabic prompts and observe unsafe generation rates below 5 percent across models. We release the dataset and code upon acceptance to support future research on dialect-aware Arabic safety evaluation. Warning: This paper contains examples of harmful and potentially offensive content included solely for research purposes.
Legal reasoning is inherently jurisdiction-dependent: the same facts can call for different legal rules and yield different conclusions across legal systems. Yet existing benchmarks rarely evaluate whether large language models (LLMs) can recognize such jurisdiction-specific variation, especially when identical fact patterns lead to divergent legal outcomes.We introduce CrossLex, a same-fact, legal-source-grounded benchmark for evaluating cross-jurisdictional legal reasoning in LLMs across three jurisdictions: China, California, and Germany. Built from authoritative legal sources, CrossLex aligns 55 legal issues spanning contract, consumer, criminal, family, and labor law, and constructs jurisdiction-aligned questions paired with answers and supporting citations. In total, CrossLex contains 6,149 instances organized into 385 fact groups, with all legal issues, answers, and cited authorities reviewed by legal professionals.To disentangle basic legal knowledge from cross-jurisdictional reasoning, CrossLex defines three complementary tasks: single-jurisdiction reasoning (T1), joint cross-jurisdictional comparison (T2), and fine-grained cross-jurisdictional evaluation (T3). We further propose Grounded Joint, a metric that jointly assesses answer correctness and legal-source grounding, and provide a unified evaluation for streamlined benchmarking. Extensive experiments on representative LLMs show that, although current models can often answer legal questions correctly, they struggle to provide accurate cross-jurisdictional legal citations.We hope that CrossLex will facilitate future research on source-grounded cross-jurisdictional legal reasoning.
RH-RAG: Trustworthy Long-Form Generation for Privacy-Constrained Settings
RH-RAG：适用于隐私受限环境的可信长文本生成
Singh, Raj Shekhar
Abstract
Generating long-form content from extensive internal reports remains challenging for organizations operating under strict privacy and security constraints, where proprietary cloud-based LLM APIs are often not viable. While locally deployed open-weight models offer a privacy-preserving alternative, existing retrieval-augmented generation (RAG) approaches on smaller models frequently lack effective global planning and accumulate factual inconsistencies over long outputs. To address these limitations, we present RH-RAG, a multi-agent framework for secure and trustworthy long form generation using local language models. RH-RAG decomposes generation into three coordinated stages: a Planner Agent that constructs a global document outline from high-level semantic summaries, a Writer Agent that incrementally generates coherent section-wise content using bounded coherence memory, and a Checker Agent that mitigates hallucinations through natural language inference-based factual verification and an attestation-driven revision loop. The framework further employs a dual-level retrieval index that supports efficient planning and fine-grained contextual generation on consumer-grade hardware. Evaluations across literary, financial, and legal domains demonstrate that RH-RAG consistently improves factual grounding, semantic coherence, and document-level alignment compared to standard and hierarchical RAG baselines, while achieving reliability competitive with proprietary cloud-based systems without compromising data privacy.
BiCAA: Bidirectional Credit Assignment for Search-Augmented Agent
BiCAA：用于搜索增强代理的双向信用分配
Huang, Yibin, Xu, Bin, Cao, Hailong, Zhu, Conghui
Abstract
Multi-step search is a fundamental capability for search agents, enabling them to iteratively acquire, refine, and integrate external evidence for complex reasoning QA. However, vanilla GRPO allocates rewards exclusively based on the model's final outputs, yielding outcome-only supervision with no supervisory signals for intermediate reasoning steps. Such sparse supervision easily causes training instability and redundant search behaviors on multi-step search tasks. To mitigate this limitation, we adopt process reward to deliver stepwise supervision signals. For this process reward, we propose two complementary criteria to judge each search step: whether the step yields new evidence to facilitate problem solving, and whether it forms an efficient, pivotal intermediate decision within the overall reasoning trajectory. Building on this insight, we propose BiCAA: a bidirectional credit assignment framework that delivers dense, distinguishing process rewards for search-augmented agents. BiCAA builds bidirectional process rewards by fusing two complementary signals: forward solvability gain and hindsight success criticality. The former quantifies step-wise improvements in answer plausibility, while the latter evaluates each step's necessity for final success via hindsight outcome-based criticality scoring. We modulate and aggregate the two signals and then fuse them with the outcome reward. Experiments on search-augmented QA benchmarks show that BiCAA stabilizes policy optimization, reduces redundant search behavior, and achieves competitive performance.
Shadow trading -- trading in a peer firm's securities on the basis of material nonpublic information (MNPI) about an "economically linked" company -- is a novel and contested theory of insider trading liability, first prosecuted in SEC v. Panuwat (2023). Enforcing it requires identifying economically linked firms ex ante, a determination the SEC makes only after the fact using mass market surveillance infrastructure. We ask whether NLP can do what the SEC's theory presumes insiders already know: identify peer firms ex ante from publicly mandated disclosures. Using a two-stage LLM pipeline applied to Item 7 (Management's Discussion and Analysis) sections of SEC 10-K filings, we score semantic similarity across 30 M&A events spanning five industries and relate similarity to announcement-day abnormal stock returns. On the Panuwat fact pattern itself the pipeline recovers Incyte among the closest peers, a sanity check on the one case with a known outcome. Across the full dataset, however, we find no association: pooling 217 peer observations, the within-event rank correlation between similarity and abnormal return is +0.07 (permutation p = 0.37), and the mean per-event Spearman correlation is +0.05 with a 95% confidence interval of [-0.08, +0.18] -- narrow enough to exclude any moderate relationship rather than merely failing to detect one. A case-level reading agrees: 14 of 30 events support the hypothesis, 12 contradict it, and 4 are ambiguous. We also find that Incyte fell outside the standard \$2B-\$10B mid-cap band on the day before the announcement, complicating the "mid-cap oncology" category the SEC invoked. These results are exploratory and bound to this pipeline, corpus, and return measure, but they put pressure on the empirical premise of shadow trading enforcement and bear on constitutional questions surrounding the SEC's financial surveillance infrastructure.
LongChart VQA: A Comprehensive Benchmark for MLLMs with Complex Multi-Chart Reasoning
LongChart VQA：针对具有复杂多图推理的多模态大语言模型的综合基准
Xiao, Ziyan, Zhu, Yinghao, Zhang, Wenting, Kim, Heaju, Yu, Lequan
Abstract
Multimodal large language models (MLLMs) are rapidly evolving with expanded context windows and stronger reasoning capabilities, enabling multi-chart understanding and multi-step inference. These abilities are increasingly important as MLLMs are adopted in complex agentic tasks. However, existing benchmarks largely emphasize single-chart perception, while simple chart-to-chart connections are insufficient to evaluate these capabilities. To capture multi-chart complexity while ensuring consistency and validity, we design a synthesis pipeline supported by latent graphs. Building on this pipeline, we introduce LongChart, a benchmark whose VQA sets contain an average of 6.5 images and 31.2 questions. We evaluate 10 state-of-the-art MLLMs and examine three factors that influence performance: reasoning patterns, auxiliary tools, and robustness to image perturbations. Our results show that MLLM accuracy decreases and varies substantially as computational complexity increases, highlighting directions for future research in multi-chart reasoning.
Prompt-Induced Waste in Large Reasoning Models: A Preregistered Two-Harness Benchmark of Coding Agents
大型推理模型中的提示引发的浪费：一项预注册的编码代理双重基准测试
Weinberger, Sarel, Hozez, Amir
Abstract
Large reasoning models used as coding agents incur costs from deliberation, tool calls, and repeated agent turns, yet the causal effect of prompt wording on this spend has not been measured systematically. We present a preregistered benchmark across six large reasoning models, two real agent harnesses, and 24 deterministic coding tasks with hidden evaluators. Across 4,643 valid runs, including screening, stress, holdout, replication, and cross-provider studies, we find that prompt formulation can multiply reasoning cost without improving correctness. Asking the model to develop and compare several approaches is the most consistently wasteful instruction, increasing reasoning tokens by 2.4-7.4x across all models. Generic "think deeply" cues also increase deliberation by 1.6-2.2x, while a bounded-efficiency template specifying scope, acceptance criteria, and a stop condition is cost-neutral and can halve reasoning. Harness choice matters even more: identical model-task-prompt triples cost 5-30x more per success under Claude Code than under pi, mainly because of larger static prefixes and more turns. Misleading architectural hints are far costlier than irrelevant prose, and provider-side caching reduces billed cost without changing behavior, so it must not be treated as efficiency. Replications on Kimi-K3 and Claude Sonnet 5 preserve the main effect directions while revealing model-specific sensitivity to thinking and certainty cues. Overall, prompt wording and harness design materially affect agent cost, often with no gain in task success.
Search-augmented large language model agents are increasingly capable of solving knowledge-intensive tasks, but their behavior when a multi-hop question is fundamentally unanswerable remains poorly understood. Existing abstention benchmarks largely expose defects at the surface of single-hop queries and therefore cannot reveal failures that emerge only after valid intermediate reasoning and retrieval. We introduce HopRefusalBench, the first controlled benchmark of refusal within multi-hop search, comprising 889 unanswerable questions constructed from KILT-grounded entity paths. It crosses three causes of unanswerability (answer unknown, false premise, and underspecified context) with root, middle, and terminal topologies, making premise verification, intermediate-bridge validation, and terminal stopping separately observable. We further propose a final-outcome taxonomy spanning target-aware refusal, pseudo-refusal, hallucinated completion, and search-budget exhaustion, together with source-aware trajectory metrics for post-trigger continuation and token waste. Across ten frontier proprietary and open-weight models in search-augmented mode, the best model achieves a target-aware correct halting rate (TCHR) of only 42.9%. Root and middle items are consistently harder than terminal items, and all models attain their highest TCHR on false premises and their lowest on underspecified questions. Yet when pooled across categories, 84.7--98.4% of each model's explicit refusal-like responses identify the correct rationale, localizing the main bottleneck to committing to an appropriate non-answer; failed trajectories instead diverge into hallucination or search-budget exhaustion. These results establish refusal in multi-hop search as a consequential evaluation problem and provide a foundation for diagnosing and improving the reliability of search-augmented agents.
Outcome-based reinforcement learning enables search-augmented language agents to learn from verifiable final answers, but its trajectory-level credit cannot distinguish the contributions of individual actions in a multi-turn search process. We propose EviSD, an evidence-conditioned self-distillation framework that uses instance-level supporting evidence as privileged information for search actions and golden answers as complementary privilege for answer actions. During training, the student samples actions from the original context, while the same model re-scores them as a privileged teacher under an action-aligned context. EviSD converts the detached teacher--student gap into a bounded correction to the outcome-derived GRPO advantage and applies it only to generated action spans. This design localizes privileged guidance while preserving the update direction determined by the outcome reward, without an auxiliary distillation objective or any change at inference time. Across seven question-answering benchmarks and three backbones spanning model scales and generations, EviSD achieves the highest macro-average Exact Match in all evaluated settings, outperforming the strongest compared methods by 1.3--2.3 points while modulating only 6.7%--15.1% of response tokens. Code is available at https://github.com/JiananXie/EviSD.
Language Equality has a Price: A Systematic Investigation of Multi-turn LLM Performance for EU-24+
语言平等的代价：对欧盟24+多轮大型语言模型性能的系统性研究
Hakimov, Sherzod, Osswald, Karl, Psurek, Jelle, Bukovszky, Eszter, Lüser, A. Altar, Schlangen, David
Abstract
We evaluate large language models (LLMs) as language agents playing goal-directed dialogue games in self-play across 30 languages: the 24 official EU languages plus six others. Unlike static or preference-based evaluation, this paradigm is multi-turn, reference-free and programmatically scored, and because the game mechanics are language-agnostic it extends to a new language by localising a fixed set of prompt and word-list files. Evaluating nine open-weight and commercial LLMs, we find that no open-weight model covers the EU-24 well: in every official language both commercial systems outscore every open-weight model, and the two weakest average below 40 points across the EU-24. The commercial systems stay ahead even in languages with four orders of magnitude less public web text, showing that linguistic parity is achievable, but not from public crawls alone. A model's home region lifts it without closing the gap: Chinese is the strongest of all 30 languages for two Chinese-developed models, yet the best Chinese score of any model belongs to a US commercial system. Coverage is also not parity of service. Pooled over models and languages, the median non-English language costs 31% more to run than English, and scores 10% lower.
When Retrieval Helps and Distracts: Evaluating Evidence-Generating LLMs for Biomedical Claim Verification
检索的帮助与干扰：评估用于生物医学声明验证的证据生成大型语言模型
Deka, Pritam, Singh, Prabhjot
Abstract
Biomedical fact-checking systems must do more than predict whether a claim is supported, contradicted, or unaddressed: they should also produce evidence that is faithful, complete, and useful for verification. We study this evidence-generation setting on CARE-XAI, a unified benchmark spanning five biomedical and health fact-checking sources. We compare base instruction LLMs, PubMed retrieval-augmented LLMs, fine-tuned LLMs, label-only LLMs, and biomedical encoder classifiers under a shared evaluation protocol. Biomedical classifiers remain strongest for verdict-only prediction, while fine-tuned LLMs are the strongest evidence-generating systems. PubMed retrieval is mixed: it helps PubMed-aligned sources such as PubMedQA and SciFact, but can distract models on broader public-health claims. We introduce Bio-GRACE, a gold-reference-normalized diagnostic for measuring whether retrieved evidence recovers the decision benefit of reference evidence. Bio-GRACE shows that retrieval utility is source-dependent, motivates selective retrieval, and exposes why retrieval recall and lexical evidence overlap are insufficient for biomedical fact-checking.
QR-Erase: Efficient Subspace-Based Machine Unlearning with Layer Localization
QR-Erase：基于子空间的高效机器遗忘方法与层级定位
Lizzo, Tyler, Heck, Larry
Abstract
Machine unlearning seeks to remove targeted information from trained models without requiring costly retraining. Existing optimization-based methods often degrade unrelated capabilities, while subspace-based approaches rely on computationally expensive singular value decompositions (SVD). We introduce QR-Erase, a subspace-based framework that uses Pivoted QR decomposition to identify and remove task-specific representations directly from model parameters. We further propose Layer-Localized QR-Erase, which restricts updates to layers containing the highest concentration of task-specific information. We show that Pivoted QR provides accurate subspace recovery with bounded error, and that under a mild spectral gap condition, the recovered subspace approaches the optimal SVD solution. Across task-level, cross-lingual, and speech unlearning, QR-Erase achieves a stronger forgetting-retention tradeoff than optimization-based methods while remaining within 5% of SVD across all metrics. Exploiting low-rank and layer-localized structure further improves forgetting (for example, reducing speech forget-set accuracy from 53.1% to 15.7%). These results demonstrate that accurate subspace recovery, rather than optimal reconstruction, is sufficient for effective unlearning and provides an efficient and general alternative to SVD-based methods for modern foundation models.
Large language models are being extensively used to simulate individual user behavior, yet faithfully representing a population requires capturing the systematic variation in values, beliefs, and cultural norms that distinguish one group from another. We introduce Population Aligned Language Models (PALMs), a suite of models each aligned to specific populations, covering five countries: USA, India, Brazil, France and Italy. PALMs are created by synthesizing rationales grounded in psychological and cultural constructs and using these as latent supervision during preference tuning for population-specific alignment. Evaluated across four dimensions: personality, values and beliefs, cultural norms, and morality, PALMs consistently outperform baselines, including culture-specialized models, achieving an average of 8.59% relative improvement over the best baseline across all five populations. Notably, construct-grounded rationales outperform both demographic prompting and survey-based fine-tuning, suggesting that grounding preference learning in psychology and culture provides a richer inductive signal than surface-level response distributions. We further demonstrate strong generalization to downstream applications with- out task-specific supervision: outperforming best baselines by 5.19% in personalized reward modeling, 6.34% in population simulation, and showing strong transfer to social reasoning tasks. Datasets and code are available at: https://github.com/limenlp/PALMs.
Chinese Translation
大型语言模型被广泛用于模拟个体用户行为，但忠实地代表一个群体需要捕捉到区分不同群体的价值观、信念和文化规范的系统性变化。我们介绍了人口对齐语言模型（Population Aligned Language Models，PALMs），这是一套针对特定人群的模型，涵盖五个国家：美国、印度、巴西、法国和意大利。PALMs 通过综合基于心理学和文化构念的理据，并在针对特定人群的偏好调整中使用这些理据作为潜在监督来创建。我们在四个维度上进行了评估：个性、价值观和信念、文化规范以及道德，PALMs 在所有五个人群中始终优于基线模型，包括文化专门化模型，平均相对提升达 8.59%。值得注意的是，基于构念的理据在性能上超越了人口统计提示和基于调查的微调，表明将偏好学习扎根于心理学和文化中提供了比表面响应分布更丰富的归纳信号。我们进一步展示了在没有任务特定监督的情况下对下游应用的强泛化能力：在个性化奖励建模中超越最佳基线 5.19%，在人口模拟中超越 6.34%，并在社会推理任务中表现出强大的迁移能力。数据集和代码可在以下网址获取：https://github.com/limenlp/PALMs。
Retrieval Augmented Biomedical Question Answering with Weak Question Recovery and Neural Reranking for BioASQ Task 14b
基于弱问题恢复与神经重排序的检索增强型生物医学问答系统——BioASQ任务14b
Zhao, Xueying, Mai, Lee, Anandganesh, Balaji
Abstract
This work presents DS@GT ARC BioASQ team's work for a biomedical question answering pipeline, integrating multi-source query expansion, neural reranking, retrieval refinement, and OpenBioLLM-assisted answer generation. The system combines PubMed retrieval with fine-tuned MiniLM-based semantic reranking, Reciprocal Rank Fusion (RRF), and feature-based relevance scoring to improve document ranking quality. To address challenging queries with weak retrieval performance, we introduce a conditional weak-question recovery strategy that applies semantic expansion, relationship-aware augmentation, and selective result merging. A post-retrieval pruning stage further removes redundant or low-relevance snippets while preserving evidence coverage for downstream answer generation. Experimental results on BioASQ evaluation batches demonstrate that the proposed recovery and cleanup strategies substantially improve retrieval robustness and MAP@10 performance on difficult question sets. The final system also incorporates output validation and post-processing steps to ensure formatting consistency and submission reliability across BioASQ phases.
Two-Stage Bengali Sentiment Classification: Domain Adaptation Through Continual Learning and Parameter-Efficient Fine-Tuning
两阶段孟加拉语情感分类：通过持续学习和参数高效微调实现领域适应
Rahman, MD Shaikh, Rabbi, Syed Maudud E, Rashid, Muhammad Mahbubur
Abstract
Understanding sentiment in low-resource languages remains a key challenge for Natural Language Processing (NLP), particularly when domain-specific data is scarce. In this work, we present SentiBanglaBERT, a two-stage Bengali sentiment classification framework combining domain-adaptive continual pretraining and parameter-efficient fine-tuning. The approach enables contextual adaptation to news-style data while remaining computationally efficient through Low-Rank Adaptation (LoRA). Beyond performance, SentiBanglaBERT integrates SHAP-based interpretability, offering linguistic insights into how Bengali morphological cues, such as negation suffixes and aspectual markers, influence sentiment predictions. Experiments demonstrate stable performance comparable to strong baselines while providing greater transparency and interpretive depth. This framework highlights the potential of domain-adaptive continual learning as a foundation for interpretable, resource-efficient NLP in morphologically rich, underrepresented languages.
Discriminative Axis, Not Data Volume: What a Contrastive Corpus Teaches an Audio Embedding
区分轴，而非数据量：对比语料库对音频嵌入的启示
Tonmoy, Abdul Basit
Abstract
Scaling the corpus is the default remedy when a contrastive representation lacks an attribute. We report a case where it does nothing, and identify what does: adding a lexical-speech round to a frozen-base multimodal embedding model raises zero-shot keyword spotting by 76 points while reducing speech-emotion recognition by 14. The loss is not a capacity limit: fine-tuning on 7,442 clips from a prosody-controlled corpus recovers emotion past its pre-speech level at a five-point keyword cost. Nor is it data volume: 29,428 mined clips whose captions explicitly name emotions, at matched exposure, move emotion by -0.0007. The difference is structural: a contrastive objective encodes an attribute only when the in-batch negatives cannot be separated without it; the controlled corpus holds sentence content fixed, so prosody is the only separating signal, whereas mined captions name emotion yet remain separable by scene content. Intervention on the same audio confirms causality: raising caption similarity does not recover emotion, but collapsing caption diversity so that emotion becomes the only separating axis recovers it by 8.9 points across three seeds, with a smaller, same-signed gain on a non-acted corpus, while keyword accuracy trades back. Corpus structure, not size or caption vocabulary, controls what a contrastive audio embedding encodes.
Answering complex questions over large document collections requires assembling complementary evidence across sections and documents. GraphRAG offers structured retrieval but typically uses fixed traversal, while agentic RAG operates over weakly structured interfaces. Our key insight is that agents should navigate document structure within and across documents rather than repeatedly search from scratch. We introduce DocNavRAG, which organizes document hierarchies and cross-region relations into a navigable graph, exposes graph operations for locating, navigating, expanding, and fetching, and maintains an evolving evidence state to guide retrieval until sufficient evidence is collected. Across four long- and multi-document QA benchmarks, DocNavRAG improves answer quality and context sufficiency over the strongest baseline by 7.8\% and 17.7\% on average.
Clinic notes and structured electronic health record (EHR) medication history often contain different medication information. Same-visit disagreement between these sources may result from note-side normalization errors, differences in terminology or timing, or actual differences in documentation. We developed a note-grounded approach that uses large language model (LLM) assisted reference construction, targeted and random human review, deterministic medication normalization, and semantic and temporal comparisons with structured medication history. We evaluated all normalization results on a patient-level held-out test set to limit adaptation to the study cohort. On 5,403 held-out mention rows, exact canonical agreement improved from 0.7226 with surface-exact matching to 0.8429 after lexical cleanup and curated alias mapping. In a random audit of previously unaudited rows, canonical-label agreement was 0.9210 among evaluable valid medication mentions, whereas treatment-action attribution was lower at 0.5326. In the full-cohort characterization analysis, only 16.44% of note-derived rows had same-visit exact overlap with structured medication history, but 55.17% had same-visit semantic overlap, 90.34% had same-visit or +/-30-day overlap, and only 3.97% remained in the strict no-structured-overlap bucket under broad project-level mapping. An ontology-backed sensitivity analysis further showed that held-out strict Observational Medical Outcomes Partnership (OMOP)-backed no-overlap fell from 43.99% to 36.68% after a development-derived alias supplement. These results show that note-to-structured-medication mismatch can arise from normalization errors, differences in terminology, and differences in documentation timing.
Semantic Alignment of AI Models: Concept Collapse, Checkpoint Dynamics, and Cross-Lingual Transfer
人工智能模型的语义对齐：概念崩溃、检查点动态与跨语言迁移
Ashoff, Tyler, Rodu, Jordan
Abstract
Language model benchmarking is a difficult task. Outcome reasoning alone does not test the model's conceptualization of language and popular open-source benchmarks are quickly saturated or ingested as training data. It is important to test the model's output, but augmenting these tests by characterizing semantic structure gives more insight to how models relate abstract concepts. However, the high dimensional embedding spaces are not easy to interpret. This work demonstrates how topological methods can be used to rigorously compare these spaces to low dimensional and interpretable baselines like ontologies and curated knowledge graphs. These multi-modal alignment tests make it possible to track model adaptations and test phrase understanding across multiple languages.
PICTURE: Enhancing Theory-of-Mind in Large Language Models by Revealing, Not Hiding, Characters' Lack of Knowledge
PICTURE：通过揭示而非隐藏角色的知识缺乏来增强大型语言模型的心智理论
Jeon, Eojin, Lee, SangKeun
Abstract
Simulating human-like Theory of Mind (ToM) has been a longstanding problem in natural language processing (NLP). To address this, existing works introduce a reasoning step of event hiding (a.k.a. perspective-taking), where events unknown to a character are removed before question answering. However, resorting to event hiding for ToM reasoning presents a performance degradation issue due to the strict output format constraints involved in event hiding. To mitigate this issue, we propose generating perspective-taking outputs as free-form explanations without event hiding, but this poses a notable yet underexplored challenge: LLMs need to inhibit responses to events unknown to characters, because the absence of event hiding exposes LLMs to these events throughout reasoning. To address this challenge, we hypothesize and empirically verify that LLMs can achieve such inhibition if a character's lack of knowledge about events is made explicit during reasoning. Based on this finding, we introduce PICTURE, a new prompting method that enables LLMs to generate a character's lack of knowledge within free-form Chain-of-Thought (CoT). Experimental results show that PICTURE outperforms existing prompting methods by an average of 7.3% on false-belief tasks.
Not the Dimension, the Norm: What Matters in Gradient-Free Weight Perturbation of Language Models
不是维度，而是范数：在无梯度权重扰动中什么才是关键
Kim, Taeyeong, Kim, Ahhyun, Kim, TaeHyeon, Lee, Unggi
Abstract
Adapting a language model to a task no longer requires training all of its weights, and a line of parameter-efficient methods has driven the trainable count from billions down to a handful of scalars. Gradient-free adaptation, which samples random weight perturbations and keeps the ones that score well, has not followed that trajectory and still perturbs every entry of the weight tensor. It is unknown whether that full-weight search is necessary, and more fundamentally which property of a perturbation makes it work at all, because existing methods vary the search space, the perturbation scale, and the aggregation together. We resolve this by intervening on one factor at a time inside a fixed pipeline, holding candidate scoring and voting constant while we vary the search dimension, the subspace that carries the perturbation, and its norm. Perturbing a frozen frame of 12 to 16 scalars stays 1.8 accuracy points behind full-weight search on average across 49 model-benchmark cells, trailing it in 36 of them. Neither the dimension nor the choice of basis explains that performance. A random frame whose Grassmann overlap with the SVD frame is at chance level performs identically once a single scale factor is matched, and at large scales the SVD directions collapse first. What survives is the perturbation norm, whose usable range closes within a factor of five across seven models and stays flat inside. The perturbation norm is therefore the one factor with a failure mode, and its safe region transfers across scale and family. The design question narrows from which subspace to perturb to how hard to shake.
Human-LLM Alignment in Language Attitudes Toward Non-Native Japanese
人类与大型语言模型在对非母语日语的语言态度中的一致性
Orita, Naho, Ogawa, Hayato, Kawahara, Daisuke
Abstract
Large language models (LLMs) increasingly evaluate human writing in high-stakes domains such as hiring and academic assessment, putting non-native speakers at particular risk. Drawing on the language attitudes framework, we compared human and LLM evaluations of parallel L1- and L2-written Japanese emails on three dimensions: fluency, status, and solidarity. Japanese raters rated L2 texts significantly lower on all three dimensions, with a fluency gap roughly twice the size of the status and solidarity gaps. Six LLM judges reproduced the direction of this bias, and five reproduced its ordering across dimensions. The models diverged from humans in two ways: all understated the solidarity gap, the most socially grounded dimension, and all differentiated among learner L1 backgrounds where humans did not. LLM judges thus reproduce native speakers' language attitudes in a structured yet attenuated form, and the language attitudes framework offers a ready-made yardstick for auditing them beyond English.
Retrieval-augmented generation (RAG) improves factuality but adds latency and engineering overhead at serving time. We propose RING (Retrieval-Internalized Generation), a holistic paradigm spanning both architecture and training that injects large-scale external knowledge into a \textit{Mixture-of-Memory Experts} and learns parametric search over this internal memory via reinforcement learning, removing the external retriever entirely. Training proceeds in three stages: continued pre-training injects new corpora into a Knowledge Expert via our novel \textit{Dual Causal Attention}; supervised fine-tuning teaches a ``search-then-answer'' pattern; and reinforcement learning with hierarchical rewards optimizes the routing-and-search policy over the parametric memory. Unlike prior parametric injection methods that pair internal memory with a fixed or rule-based retriever, RING {learns} its retrieval policy directly from task signals. We further frame RING theoretically as a search-free approximation to the classical RAG objective. To evaluate large-scale injection of genuinely {new} knowledge without test-time leakage, we further construct News-2025, a benchmark built from news strictly post-dating the base LLM's pretraining cutoff. RING matches or surpasses both search-based RAG and parametric injection baselines in accuracy and efficiency.
Does Accuracy Equal Evidence? Reasoning Faithfulness under KV Cache Compression
准确性等于证据吗？KV缓存压缩下的推理可信性
Ai, Mengting, He, Jingrui, Guo, Yue
Abstract
KV cache compression is commonly evaluated by final-answer accuracy, implicitly assuming that preserving the answer also preserves the reasoning that supports it. We test this assumption for large reasoning models and show that it can fail: under compression, correct answers and the validity of their visible supporting rationales can be preserved at different rates. We study this failure with a controlled fixed-trace replay protocol, which holds reasoning content fixed and isolates whether compression preserves usable information from an already available trace. We evaluate ten token-eviction KV compression methods and one quantization method on three models across mathematical reasoning, scientific QA, clinical calculation, and long-context retrieval. We measure final accuracy, answer-chain consistency, and perturbation faithfulness. Across tasks, token-eviction methods can preserve competitive final-answer accuracy while substantially degrading chain support or perturbation faithfulness. We call this the answer-evidence gap. A coverage-preserving quantization control is substantially less affected, suggesting that the failure is tied less to KV memory reduction itself than to losing access to parts of the reasoning trace. Code is available at https://github.com/famous-blue-raincoat/Safe_KV_Compress.
However, whether these judges truly evaluate the scientific substance of ideas or are influenced by superficial stylistic presentation remains an open question. To address this question, we propose SciStyleBench, a unified three-component benchmark for diagnosing and mitigating stylistic bias in LLM-based idea evaluation: (i) First, SciStyleStage, a three-stage evaluation environment that applies controlled stylistic perturbations to fixed scientific content across three settings no context, fixed-domain context, and open-domain retrieval context, covering 600 scientific ideas and 15 style variants, with 9,000 evaluation instances per setting; (ii) Second, SciStyleMetrics, a set of quantitative measures, including Style Bias Index (SBI), Substance Recognition Rate (SRR), and Adversarial Win Rate (AWR), to characterize how stylistic variation affects scoring stability, substance discrimination, and ranking robustness; (iii) Third, SciStyleExtractor, a plug-and-play evaluation module that separates presentation style from scientific content by predicting style type and deviation before style-conditioned evaluation, enabling us to assess whether style awareness reduces stylistic bias. Experiments on SciStyleBench show that direct LLM judges remain sensitive to writing style and struggle to distinguish scientific substance. In contrast, SciStyleExtractor reduces SBI from 0.566 to 0.501 while increasing SRR and AWR from 0.504 and 0.554 to 0.759 and 0.899, respectively. These results suggest that robust idea evaluation requires invariance to stylistic variation without sacrificing sensitivity to scientific substance. Overall, SciStyleBench provides a systematic framework for identifying, quantifying, and mitigating stylistic bias in scientific idea evaluation.
Learning What to Remember: Test-Time Training via Context Distillation
学习记忆的内容：通过上下文蒸馏进行测试时训练
Wang, Zixuan, Dang, Xingyu, Zhu, Rui-Jie, Wen, Zixin, Fu, Hengyu, Chai, Wenhao, Lee, Jason D.
Abstract
Effective long-context modeling is not merely about retaining more of the past, but about preserving the information that may prove relevant later. Test-time training (TTT) is an appealing approach that performs online parameter updates for long-context modeling, yet existing TTT methods only optimize either reconstruction or online adaptation objectives without considering the future utility of retained information. In this work, we propose \textbf{T}est-\textbf{T}ime \textbf{C}ontext \textbf{D}istillation (TTCD), a TTT framework that introduces a self-supervised objective for allocating limited memory capacity for future use. Specifically, TTCD uses a long-window teacher to supervise the fast weights of a short-window student, where the hidden-state discrepancy between them offers a dense, self-supervised signal guiding the model to memorize the contextual information crucial for future token predictions. We focus on an in-place variant: In-Place TTCD (IP-TTCD), which uses the existing MLP parameters as the fast weights. Experiments on long-context language modeling tasks show IP-TTCD consistently outperforms DeltaNet, Gated DeltaNet, sliding-window attention, and TTT when pre-trained from scratch. Furthermore, IP-TTCD allows pre-trained transformer models to adapt their parameters during inference through continual pre-training, gaining long-context capabilities with only a lightweight architectural augmentation. Our results position TTCD as a step toward architectural continual learning.
Sparse attention is widely deployed in long-context serving stacks, yet no framework audits how discarding blocks changes the influence of specific content on model output. We first establish that the phenomenon is real and causal: Block Sparse Flash Attention (BSFA) route replay across four architectures changes output decisions in 13 of 16 cells, with zero identity-replay label flips. We then introduce a dense-calibrated counterfactual audit using matched probe cards---Gold (carrying the correct answer label), Poison (carrying a target wrong label), and Benign (filler only)---under six-layout position symmetry, isolating the sparsification-specific effect. Two patterns compete. Signal concentration: the selector preserves Gold and Poison blocks far above filler-matched Benign blocks (G$\approx$P$\gg$B across all model--task pairs). Integration loss: discarding blocks severs cross-block attention---confirmed by an ablation where isolating the probe block collapses its influence from 4.48 logits to zero. Compression ratio governs the balance: a full sweep from mild ($c=0.25$) to aggressive ($c=0.75$) compression across four model--task pairs reveals that three of four cells move toward stronger sparse amplification at higher compression, with two exhibiting sign reversals. Three independent arms---BSFA route replay, controlled block-top-$k$, and KV-cache eviction---converge: sparsification changes content influence in ways aggregate accuracy cannot detect. We provide an open measurement framework deployable on any model exposing block identities.
PGMem: Tightly Coupled Persona-Memory Graph for Lifelong Personalized Agents
PGMem：紧密耦合的人格-记忆图用于终身个性化智能体
Choi, Wonjun, Kim, Yerim, Lee, Yukyung, Yoon, Susik
Abstract
Long-term personalized dialogue agents must track user preferences as their personas evolve. Existing memory systems organize past events well, but store personas as flat profiles detached from the events that justify them. This loose coupling leads to the memory-persona validity gap and the persona-aware retrieval gap. We propose PGMem, a heterogeneous persona-memory graph that connects event and persona nodes through typed provenance and evidence edges, keeping each persona signal traceable to the events that support or revise it. At retrieval time, PGMem expands from query-relevant seeds and ranks signals by evidential validity. Across three benchmarks with small language model backbones, PGMem consistently outperforms summary-based, persona-aware, graph-structured, and agentic memory baselines, and improves performance as the context grows. The source code of PGMem is available at https://github.com/wonjunchoi23/pgmem/
TIDES: A Longitudinal Bilingual Dataset for Modeling Multi-Party Social Dynamics
TIDES：用于建模多方社会动态的纵向双语数据集
Lee, Heechan, Kang, Jeonggyu, Myung, Junho, Jeong, Jaywoong, Kim, Juho, Seering, Joseph
Abstract
Group conversations are fundamental to human collaboration, yet standard large language models (LLMs) still struggle with the complexities of multi-party interaction. This challenge persists in part because existing group conversation datasets are often limited to short-term lab settings with contrived tasks, failing to capture the long-term social dynamics of real-world teams. To bridge this gap, we introduce TIDES, a high-resolution longitudinal dataset tracking 12 university project teams over a full semester. Comprising 75,971 utterances in both English and Korean from in-person meetings, TIDES provides a naturalistic record of teams working on self-managed projects. Our socio-structural annotations-covering interaction types, emergent roles, and development stages-allow for modeling of team evolution over months. Experiments show that fine-tuning on TIDES improves next-speaker prediction by 13.8 percentage points over a bigram baseline (64.53%) and yields performance comparable to strong proprietary zero-shot models. The model also comes within 2.1 percentage points of the published state of the art on the AMI Meeting Corpus while using approximately 42% less training data. However, human evaluations suggest that better next-speaker prediction does not necessarily yield more natural or coherent utterances, as fine-tuned models were generally less preferred than vanilla models. This potential mismatch motivates further study of how structural modeling can support natural multi-party generation.
Constructing Parallel Multidimensional Chromatic Lexicons for Corpus-Assisted Analysis of Russian and English Texts
构建平行多维色彩词典以辅助分析俄语和英语文本
Nikitina, Larisa
Abstract
This article addresses the relative scarcity of research tools for the corpus-assisted linguistic analysis of colour terms in literary texts. It describes the development of two multidimensional chromatic lexicons: one for Russian (224 entries) and one for English (141 entries). Lexicon construction involved sourcing colour vocabulary from specialised resources and research literature, comparing the two language inventories, manually checking translated candidates, and addressing language-specific morphological features. In addition to identifying colour terms and visual descriptors, the lexicons classify entries according to hue, saturation, and temperature. To demonstrate their practical application, a pilot study was conducted on purposively sampled corpora of poetry by Andrei Bely (20,373 tokens) and Emily Dickinson (28,479 tokens). All retrieved matches were checked in context and classified as Confirmed_chromatic, Ambiguous_visual, or Excluded. The analysis was implemented in two main stages: a strict analysis including confirmed chromatic lexis only, followed by a sensitivity analysis incorporating both confirmed and ambiguous chromatic lexis to determine whether coding decisions about borderline cases affected the main findings. The quantitative results indicated marked differences in the use of colour terms, visual descriptors, hue, saturation, and temperature. Specifically, the analysis revealed that confirmed chromatic terms occurred 3.4 times more frequently in the sampled Bely corpus than in the Dickinson corpus. These findings demonstrate the analytical value of a multidimensional approach, with the main contribution of this study being a transparent and reusable procedure for constructing and applying multilingual chromatic lexicons.
RADAR: Rubric-Aware Dependency and Redundancy Analysis for LLM-as-Judge Evaluation
RADAR：基于评分标准的依赖性和冗余性分析用于LLM作为评审的评估
Singh, Divyansh, Davari, Reza, Mashhadi, Afra
Abstract
Rubric-based LLM-as-judge pipelines often assume that evaluation criteria provide independent signals. In practice, however, criteria can be behaviorally coupled: improving one criterion may systematically change scores on another, distorting aggregate scores used in model-release or product-update decisions. We introduce RADAR, a lightweight preflight diagnostic framework for estimating such coupling before large-scale evaluation. Given a rubric, RADAR generates targeted synthetic probes, scores each probe on all criteria, and produces a directional coupling matrix that shows which criteria co-score and how. We validate RADAR on three industry-relevant evaluation settings: NVIDIA HelpSteer2, SumPubMed, and the Yale-Salesforce SummEval benchmark. Using only a small number of probes per criterion, RADAR recovers human inter-criterion correlation structure (Pearson r > 0.84) and provides practitioners with concrete audit signals about redundancy, hierarchy, and aggregation sensitivity before committing to large-scale judging.
Chinese Translation
基于评分标准的LLM作为评审的流程通常假设评估标准提供独立的信号。然而，在实践中，这些标准可能在行为上是相互关联的：改善一个标准可能会系统性地改变另一个标准的评分，从而扭曲用于模型发布或产品更新决策的总评分。我们提出了RADAR，这是一个轻量级的预检诊断框架，用于在大规模评估之前估计这种关联性。给定一个评分标准，RADAR生成针对性的合成探针，对每个探针在所有标准上进行评分，并生成一个方向性关联矩阵，显示哪些标准共同评分以及如何共同评分。我们在三个与行业相关的评估设置上验证了RADAR：NVIDIA HelpSteer2、SumPubMed和Yale-Salesforce SummEval基准。使用每个标准仅少量探针，RADAR恢复了人类标准间的相关性结构（Pearson r > 0.84），并在进行大规模评审之前为从业者提供了关于冗余性、层次结构和聚合敏感性的具体审计信号。
Divergent large language model predictions from convergent representations in ambiguous word pairs
模糊词对中收敛表示的发散大型语言模型预测
Scott, K. Jack, Pat, Narun, Liesaputra, Veronica
Abstract
In this work we investigate how decoder-only transformers resolve lexical ambiguity through layer-by-layer analysis of three models spanning three parameter sizes (GPT-2-Small-117M, Llama-3.2-3B, Qwen2.5-32B). For both homonyms and polysemes, we find that representations become maximally distinct in middle layers, then partially reconverge in late layers, while the KL divergence between their next-token predictions reaches its maximum in the final layers. The activation patching experiment provides causal evidence that late-layer representational differences directly determine outputs despite apparent increased similarity in embedding space. Our single-layer ablation experiment indicates that models achieve equivalent disambiguation despite qualitatively different layer-wise vulnerabilities. These findings offer a mechanism for recent observations where models' internal embedding similarities show low correlation with their behavioural outputs despite strong performance. The semantic distinctions therefore remain present but become increasingly invisible to similarity measures over the embeddings, with implications for embedding-based methods such as semantic search, retrieval, and clustering that rely on late-layer cosine similarity.
Analyzing Speech Condition Effects in Dysarthric ASR: A Layer-wise Probing Study
分析运动性语言障碍自动语音识别中的语音条件影响：逐层探测研究
Muthu, Darwin Jelestin, Gupta, Navya, Tay, Wei Lin, Zhang, Zhengchen, Zhengkui, Daniel Wang, Tong, Rong
Abstract
Automatic speech recognition (ASR) performance degrades sharply on dysarthric speech, yet how disordered articulation reshapes a model's internal representations is underexplored. We present a layer-wise probing analysis of a transformer ASR encoder on Mandarin dysarthric speech under three transcript-matched conditions: original dysarthric speech, speaker conditioned zero-shot TTS resynthesis, and unconditioned TTS. The probes reveal a task-dependent hierarchy: phoneme boundary information stays weak for dysarthric speech at every layer, phoneme identity becomes recoverable toward the upper layers, and recognition difficulty is encoded in the deepest layers. Tone-sensitive evaluation shows Mandarin lexical tone is a persistent error source. Cross-condition similarity divergence grows with depth, indicating that disordered speech affects high-level representations more than low-level acoustic features. Guided by these findings, single-layer LoRA at layer 7 and adaptation on subset layers 5-8 achieve performance within 3.5% and 2.48% relative margins of full encoder adaptation, respectively, while upper-layer adaptation is less effective for dysarthric speech. These findings link representation analysis to parameter-efficient fine-tuning and motivate layer-aware adaptation for low-resource Mandarin dysarthric ASR.
Large language models (LLMs) are increasingly extended into deep search agents that solve complex questions through multi-step interaction with external search and browsing tools. However, existing agents often incur substantial computational and interaction costs, generating lengthy trajectories that contain redundant queries, inefficient exploration, and irrelevant observations. Existing efficiency-oriented methods usually encourage agents to use tools less frequently, but treating all tool interactions uniformly may also suppress steps that gather necessary evidence. In this paper, we propose CRISP, a framework for training efficient deep search agents through critical step perception. Unlike prior efficiency methods that uniformly penalize tool use, CRISP distinguishes interactions that gather necessary evidence from redundant ones and shapes the training reward to preserve the former while pruning the latter, improving efficiency without sacrificing the evidence needed for correct answers. Specifically, CRISP first constructs critical-step labels with Backward Evidence Induction: starting from the final answer, a strong model traverses a completed search trajectory backward and judges whether each tool-interaction step provides or preserves evidence for the final answer. We then distill these step-wise judgments into a smaller critical-step recognizer, enabling full-trajectory analysis in a single pass. During policy optimization, an efficiency-aware reward is applied only to successful rollouts. Experiments on BrowseComp and HLE-Verified show that CRISP maintains competitive final-answer accuracy while reducing average interaction turns by 15.1% and 33.2%, respectively, demonstrating substantial improvements in interaction efficiency.
Multimodal Large Reasoning Models (MLRMs) have achieved strong performance on tasks requiring visual understanding and multi-step inference. However, as reasoning trajectories grow, models may become less effective at using information established earlier in the context, increasing the risk of reasoning errors. Existing approaches primarily address this problem by sustaining visual grounding throughout reasoning. However, reasoning also transforms visual observations into task-specific relations, constraints, and intermediate conclusions whose influence may weaken over long trajectories. Our attribution analysis suggests that correctness is not consistently separated by image attribution alone, but is more closely associated with whether trajectories retain and integrate such reasoning-derived information across stages. Motivated by this, we introduce TRAM (TRajectory-derived Auxiliary Memory), a training-free method that augments standard decoding with an auxiliary memory pathway derived from the model's own reasoning trajectory. TRAM consolidates completed reasoning into a compact latent memory, updates it online through fast and slow recurrent streams, and feeds it back into selected decoder layers through a lightweight residual pathway. Experiments across four MLRM variants on eight benchmarks show that TRAM improves performance over vanilla decoding on mathematical, scientific, and general visual reasoning tasks without additional training.
Automatic Annotation of Ancient Greek Vowel Length
古希腊元音长度的自动标注
Cleland, Albin Thörn, Cullhed, Eric
Abstract
Prior work in Ancient Greek NLP relies on corpora that do not disambiguate the phonemic vowel length of alpha, iota, and ypsilon, together known as the dichrona. Depending on lexeme, morphology, sandhi, syntax, and conventions of period, genre, and verse form, each of these letters can represent either a long or a short vowel. Deciding and marking the correct length is known as "macronizing", a long-tail problem given the sheer mass of word forms and the context dependency of individual instances. No macronized corpus of Ancient Greek is publicly available at scale, so a stand-alone macronizer is needed. While previous work has shown how to build a static, corpus-bespoke vowel-length dictionary, the present paper constructs the first general-purpose macronizer for arbitrary Ancient Greek input. Given input carrying lemma, part-of-speech, and morphological annotation in the standard CoNLL-U format, a set of recursive modules lets less common word forms inherit markup from more common forms of the same lexical word. The macronizer's chief application is generating training data for machine learning: we show that a small character-level transformer trained on the macronizer's own output learns to generalize past the cases the rule-based system leaves unmarked, matching or exceeding its accuracy on a gold-standard, manually annotated benchmark of verse and prose. We also show that macronization can improve downstream prosodical NLP tasks like verse scansion.
On-policy distillation (OPD) provides teacher supervision on states visited by the student, reducing the distribution gap between training and inference. However, in multi-turn agentic tasks, student deviations may accumulate over time, gradually moving the trajectory away from states where teacher guidance remains effective. Our quantitative analysis further shows that high-disagreement states offer promising opportunities for teacher guidance, but determining whether such guidance is beneficial requires examining its effect on subsequent student trajectories. We propose FutureBridge-OPD (FTB), which executes a short teacher bridge at a high disagreement state and uses the resulting student continuation to assess whether the bridge increases the density of positive distillation signals relative to the teacher. On ALFWorld, WebShop, and ScienceWorld, under the main Qwen3-32B teacher to Qwen3-1.7B student setting, FTB outperforms vanilla OPD and TCOD by an average of 16.6 and 7.6 points, respectively, and remains effective across student scales and teacher settings. Our code is publicly available at https://github.com/ChenChiShui/FutureBridge-OPD.
LLM companions are deployed at scale in personally consequential settings, yet poorly evaluated. Existing benchmarks use hand-authored scenarios and prompted simulators, aggregate empathy into one score, and overlook judge biases such as same-family favoritism and scale drift. We introduce CompanionBench, an interactive bilingual benchmark. To our knowledge, it is the first companion benchmark to ground both its scenarios and a trained user simulator in de-identified real-world data. A hidden disclosure gate branches each persona's trajectory on the agent's own behavior, controlling the interaction state space without scripting dialogue. We operationalize ten capabilities derived from 25 theories across psychology and counseling, four of them not graded explicitly by prior work: holding ambiguity, selfobject responsiveness, positive resonance and calibrated challenge. Agents are assessed on two complementary axes: a subjective ten-capability rubric and a deterministic measure of whether deeper disclosure was earned. A cross-family panel dilutes same-family favoritism; an Item Response Theory model separates agent quality from judge severity. Theory fixes what to measure and how personas are structured; real data supply events, history, and profiles -- coverage from theory, authenticity from data. Rankings are reproducible in both languages (rho = 0.996 ZH / 0.953 EN). Evaluating 28 agents reveals capability-level differences obscured by aggregate scores. Emotion regulation and calibrated challenge remain common weaknesses; holding ambiguity discriminates most. Role-play agents rank near the bottom: immersion does not imply relational competence. Across agents, the dominant failure mode is substituting surface warmth for substantive relational support. We will release 500 Chinese-English parallel pairs and the evaluation code.
Can a strictly local, iterated, weight-shared computation primitive support language modelling, and which of those three properties actually drives the model's behaviour? We define \textsc{TextNCA}, a 1D causal windowed-attention realisation of the Neural Cellular Automaton primitive, and study a hierarchical variant that cascades three stages with windows $w \in \{8, 32, 128\}$ and $T_s$ shared-weight iterations per stage, all on WikiText-103 at roughly 30M parameters and 60k training steps. The model does not match a parameter-matched Transformer at this scale (Hier-TextNCA $60.3$ vs.\ Transformer-6L $52.8$ and Transformer-12L $44.7$ PPL), so we treat it as an analytical probe rather than a proposed alternative. The behaviour we observe is largely explained by the staged narrow-to-wide schedule: a non-iterating sliding-window Transformer that reuses the same schedule comes within $+4.1$ PPL of the iterated model, while reversing, flattening, or breaking the monotonic ordering of the schedule costs between $+16.7$ and $+70.8$ PPL. Iteration adds a smaller bounded benefit on top of the schedule, with a clear optimum at $T_s{=}4$ and a U-shaped degradation beyond it. The GRU gate and learned per-step embeddings are required for that benefit to appear, and training with random $T_s$ yields an inference-time iteration-count knob at the cost of substantially higher absolute PPL. We position the work as a controlled reading of which parts of NCA-style computation carry the weight in language modelling.
Large vision-language models (LVLMs) have achieved substantial performance gains in Video Temporal Grounding (VTG) through reinforcement learning (RL). However, existing methods primarily rely on outcome correctness rewards that evaluate only the final predicted intervals, leaving boundary-related visual evidence and its correspondence with timestamp predictions insufficiently constrained. In this paper, we delve into timestamp prediction and its underlying boundary-level visual evidence, showing prevalent misalignment between visual evidence and predicted timestamps across widely used benchmarks. To address this issue, we propose Competence-Aware Visual Boundary Evidence Alignment (CAVE), which augments localization optimization with boundary-specific visual evidence rewards to mitigate evidence-timestamp misalignment. Specifically, to explicitly represent the boundary-specific visual evidence, CAVE introduces boundary-specific evidence tokens and initializes their structured generation and distinct boundary semantics through a lightweight supervised warm-up. During RL, the visual boundary evidence alignment reward reinforces the visual attention of special evidence tokens within the ground-truth boundaries, thereby promoting alignment between visual evidence and temporal boundaries. Moreover, performance-aware gating for evidence supervision is designed to adaptively retain evidence guidance for poorly localized groups while reducing it once localization becomes sufficiently accurate to avoid over-constraining fine-grained boundary refinement. Extensive experiments on several public VTG benchmarks demonstrate the effectiveness of our method.
Recent advances in Reinforcement Learning (RL) have substantially improved the capabilities of autonomous search agents, enabling sophisticated planning, and iterative retrieval over dynamic information sources. However, optimizing language models for specialized search behaviors often incurs an alignment tax, where gains in search performance come at the expense of general-purpose capabilities, limiting their effectiveness as universal assistants. In this technical report, we present the training framework behind the Yuanbao search agent, designed to achieve search specialization without sacrificing general intelligence. Built upon the Hunyuan3 architecture, our framework combines agentic reinforcement learning for autonomous search with a cross-domain expert On-Policy Distillation (OPD) pipeline. Experts specializing in complementary general-purpose domains are distilled into the search-specialized student, restoring and further enhancing its broad capabilities. Rather than treating specialization and general capability as competing objectives, our hybrid training strategy jointly optimizes both, effectively mitigating the alignment tax. Extensive experiments demonstrate that the resulting model achieves competitive search performance while consistently improving its general-purpose capabilities, providing a favorable balance between specialized execution and broad generalization in real-world search scenarios.
Executing long-horizon tool invocations in real-world environments is severely challenged by dynamic user intent noise. Existing methods attempt robustness via implicit history scanning or text compression, yet predominantly assume perfect instructions in simplistic scenarios. Inevitably, under fluctuating contexts, obsolete constraints dilute model attention, triggering catastrophic intent deviation and infinite API loops. To resolve this, we propose IACM-RL, a comprehensive framework for robust tool invocation. First, we introduce the DynamicIntent pipeline, synthesizing trajectories across 13 fine-grained fluctuation scenarios, paired with a five-dimensional diagnostic metric suite. Second, IACM-RL deploys a BeliefState-based Self-Generated Context Manager that proactively tracks shifting goals and isolates overwritten parameters using structural stale flags. To autonomously internalize this state-tracking capability, we optimize the policy using a hierarchical intent-driven reward alongside three auxiliary losses (action calibration, CM extraction, and state distillation). Experiments on DynamicIntent, BFCL-V3, and $\mathrm{\tau}^2$-Bench demonstrate that IACM-RL significantly outperforms baselines, reducing infinite loops and stale context errors while enhancing out-of-domain generalization.
Speculative decoding accelerates LLM inference only when drafted continuations survive target-model verification. Semi-autoregressive drafters such as DSpark predict an entire token block with one backbone forward and refine it with a lightweight Markov head. However, DSpark decodes this block as a single chain, so an early mismatch invalidates the remaining suffix and limits the benefit of large draft blocks. We show that the conditional structure already learned by DSpark can support multiple parent-consistent continuations without retraining or additional backbone passes. We introduce Parent-Conditioned Drafting Tree (PCTree), which uses the pretrained Markov head to score alternative children separately for each concrete parent and allocates a fixed verification budget to the most probable paths. This converts DSpark's linear draft into a tree while preserving its one-pass parallel backbone. Across Qwen3-{4B,8B,14B} and nine benchmarks, at $B{=}7$, measured speedup gains over autoregressive (AR) decoding, relative to matched DSpark, range from $3.1\%$ to $29.5\%$. On Qwen3-4B GSM8K at $B{=}16$, PCTree increases mean acceptance length from $9.41$ to $11.16$ and three-run mean AR speedup from $6.14{\times}$ to $6.60{\times}$. These show that parent-conditioned branching can turn conditional capacity already present in a semi-autoregressive drafter into end-to-end inference gains through an inference-only change.
Current speech translation systems, including SpeechLLMs, are trained on cleaned text and tend to strip disfluencies like filled pauses and false starts rather than translate them. We show this comes at a cost: disfluencies carry meaning that gets lost when speech is cleaned up. To study this systematically, we introduce Uh-Mazing, a benchmark of human-translated, disfluency-annotated Switchboard speech covering English into eight target languages. Across these languages and several architectures, we find that false starts and self-repairs, not filled pauses or discourse markers, drive most of the translation-quality loss, and that models which fail to preserve a disfluency tend to omit it rather than mistranslate it. We show inference-time decoding can mitigate this without retraining, and release the benchmark and code.
Large language models (LLMs) capable of self-improvement require not only effective policy optimization, but also a principled mechanism for transforming transient interaction experience into persistent model capabilities. Existing self-improvement paradigms remain fragmented: test-time methods can explicitly extract experience but cannot internalize it into model parameters, whereas training-time optimization methods can update model parameters but lack an explicit mechanism for accumulating transferable experience. Bridging these two paradigms requires a critical intermediate stage that remains underexplored, namely \emph{experience distillation}. To address this gap, we propose \textbf{SPEE} (\textbf{S}elf-\textbf{P}rogressive \textbf{E}xperience \textbf{E}volution), a unified post-training framework that sequentially performs explicit experience evolution followed by implicit policy optimization. During explicit experience evolution, SPEE reflects on trajectories collected from multiple interactions to extract, verify, and progressively evolve transferable experience, which is subsequently internalized into the policy through privilege-guided On-Policy Self-Distillation (OPSD). During implicit policy optimization, reward-driven reinforcement learning leverages these internalized priors to explore novel solution strategies. In the experience evolution stage, a continuously evolving global experience pool consolidates knowledge from both successful and failed trajectories, filters out low-utility experience, and mitigates post-hoc rationalization induced by individual trajectories. Experiments on five mathematical reasoning benchmarks demonstrate that SPEE consistently outperforms both test-time and training-time self-evolution baselines across three model scales. The source code is available at https://github.com/rrrsj/SPEE.
Recent advances in neural text-to-speech (TTS) systems have substantially improved speech naturalness and intelligibility across many languages. However, comprehensive evaluation methodologies that jointly assess perceptual quality, speaker similarity, and acoustic fidelity across diverse speech domains remain limited, particularly for low-resource and underrepresented languages. This paper presents a reproducible, multi-metric benchmarking framework for systematic evaluation of modern TTS systems through domain-specific analysis. The proposed framework integrates complementary subjective and objective evaluation protocols and is demonstrated through a comprehensive case study on a representative low-resource language spanning four speech domains: Formal, Conversational, Literary/Storytelling, and Emotional. Four state-of-the-art TTS systems -- Indic-Parler-TTS, MMS-TTS, Microsoft Edge TTS, and Google Gemini TTS -- are evaluated using MUSHRA listening tests, ABX discrimination tests, speaker similarity scoring with Resemblyzer, and acoustic analyses based on mel-cepstral distortion (MCD) and F0 RMSE over 960 audio pairs. Results reveal substantial variation in TTS performance across speech domains, with emotional speech consistently presenting the greatest synthesis challenge (mean MCD 12.03 dB; mean F0 RMSE 889 cents), while conversational speech achieves the highest overall acoustic fidelity. Beyond the empirical findings, this work provides a reproducible evaluation framework, publicly releasing evaluation scripts, result tables, and executable Colab notebooks to support standardized benchmarking and future research on TTS evaluation for low-resource languages.
The rapid spread of false and misleading health information through digital platforms has become a major public health challenge, particularly during infectious disease outbreaks where delayed verification can influence public behaviour and hinder effective disease control. Although recent advances in automated health misinformation detection have shown encouraging results, most existing approaches rely heavily on global biomedical resources and often fail to capture the local context needed to verify claims in developing countries. This study presents a retrieval-augmented transformer framework designed to verify health-related claims using trusted evidence from the World Health Organization and the Nigeria Centre for Disease Control and Prevention. The framework combines semantic evidence retrieval with transformer-based classification to determine whether a claim is true, false, or misleading. To evaluate the proposed approach, a manually annotated dataset of 67 verified health claims covering coronavirus disease, Lassa fever, cholera, measles, and monkeypox was compiled from Nigerian fact-checking sources. Three transformer models and a retrieval-augmented configuration were evaluated. The Bidirectional Encoder Representations from Transformers model achieved the best performance, with an accuracy of 71% and a weighted F1-score of 0.66. Although retrieval augmentation did not improve classification performance because the current evidence repository was limited in size and coverage, the findings highlight the importance of comprehensive and authoritative knowledge sources for reliable health misinformation verification. The proposed framework provides a practical foundation for developing context-aware and evidence-driven health misinformation verification systems for Nigeria and other resource-constrained settings.
Chinese Translation
通过数字平台快速传播的虚假和误导性健康信息已成为一个主要的公共卫生挑战，特别是在传染病暴发期间，延迟验证可能影响公众行为并阻碍有效的疾病控制。尽管近期在自动化健康虚假信息检测方面取得了令人鼓舞的进展，但大多数现有方法严重依赖全球生物医学资源，往往无法捕捉到验证发展中国家声明所需的本地背景。本研究提出了一种检索增强型变换器框架，旨在利用世界卫生组织和尼日利亚疾病控制与预防中心的可信证据来验证与健康相关的声明。该框架结合了语义证据检索与基于变换器的分类，以确定声明是真实的、虚假的还是误导性的。为了评估所提方法，编制了一个手动注释的数据集，其中包含67个经过验证的健康声明，涵盖了冠状病毒病、拉沙热、霍乱、麻疹和猴痘，数据来自尼日利亚的事实核查来源。评估了三种变换器模型和一种检索增强配置。双向编码器表示变换器（Bidirectional Encoder Representations from Transformers）模型表现最佳，准确率为71%，加权F1-score为0.66。尽管检索增强未能改善分类性能，因为当前的证据库在规模和覆盖范围上有限，但研究结果强调了全面和权威的知识来源在可靠健康虚假信息验证中的重要性。所提出的框架为开发适应本地背景和以证据为驱动的健康虚假信息验证系统提供了实用基础，适用于尼日利亚及其他资源有限的环境。
Can AI Agents Simulate A/B Test Outcomes? A Validation Framework for Agentic Experimentation
人工智能代理能否模拟A/B测试结果？一种代理实验的验证框架
Hut, Stefan, Masoero, Lorenzo
Abstract
A/B testing remains the standard for rolling out new features in the technology industry. Each experiment, however, consumes real traffic, engineering effort, and weeks of wall-clock time. Can AI agents---conditioned on behavioral profiles and contextual descriptions of the intervention---simulate outcomes accurately enough to vet candidate treatments before committing live traffic? We formalize this question as a \emph{Simulated Randomized Controlled Trial} (S-RCT) and derive a two-layer error decomposition that separates agent approximation error from subsampling error, enabling targeted improvements to each. The framework is agent-agnostic: any behavioral model---from a fine-tuned specialist to a general-purpose foundation model---can serve as the simulation engine. Validated on 67 historical marketing A/B tests, a baseline S-RCT using an off-the-shelf foundation model captures directional signal (sign overlap 0.70) but systematically overshoots effect magnitudes. A two-phase pre-period calibration protocol reduces the squared prediction error (after removing irreducible measurement noise) by ${\sim}77\times$; a within-subject design---where each agent is exposed to both arms---reduces standard errors by ${\sim}2.4\times$. We discuss limitations of the current approach and identify applications where experimenters stand to benefit from agentic signals.
Global Optimization and Inference-Time Region Grafting for Agentic Workflows
代理工作流的全局优化与推理时区域嫁接
Koh, Donghyeok, Kim, Gyuwan, Bak, Jinyeong, Na, Seung-Hoon, Yang, Tao, Jang, Haneol, Park, Cheoneum
Abstract
Recent advances in agentic workflow optimization automate workflow design through task-specific workflow search or input-conditioned architecture selection. However, they determine the workflow before execution and cannot adapt failed workflow regions using execution-time label-free quality signals. Naively enabling such inference-time adaptation through whole-workflow re-optimization would be computationally prohibitive. To tackle this challenge, we introduce GRAFT, which preserves a globally optimized workflow while locally replacing only selected regions for each input. Without parameter training, GRAFT evaluates region-level alternatives using label-free execution-quality signals and accepts only replacements that improve local quality while preserving workflow-level consistency, thereby enabling instance-wise adaptation without whole-workflow re-optimization. GRAFT applies without modification across a range of tasks spanning mathematical reasoning, code generation, and multi-hop and knowledge-intensive question answering. Under matched optimizer and executor settings, it improves over the strongest prior workflow-optimization method, MaAS, by 3.85 points on average. Replacing only the executor with a stronger model yields further gains without re-optimizing the global workflow. This suggests that an optimized workflow is not merely a static optimization artifact, but an adaptable execution policy that can evolve with inference-time feedback and stronger executors.
To operate robustly in open-world environments, autonomous agents should be able to infer the behavior of unfamiliar systems through interaction alone, even in the absence of documentation. However, existing tool-use benchmarks expose semantic tool schemas in static environments, allowing agents to rely on prior knowledge rather than autonomous discovery. To address this limitation, we introduce ScrambleToolBench, an interactive terminal benchmark designed to isolate behavioral reasoning. By removing semantic cues and enforcing a continuous task curriculum, the benchmark requires agents to uncover hidden tool behaviors entirely through trial-and-error interaction. The benchmark further introduces dynamic challenges, including mapping drift, stochastic action failures, and temporal execution windows, to evaluate whether agents can revise and adapt their hypotheses as the environment changes. Our evaluation of state-of-the-art language models reveals that successful initial discovery does not translate into robust adaptation. When faced with structural changes such as mapping drift, agents fail to use deductive strategies such as cycle tracing, and instead exhibit belief inertia or fall back to exhaustive search. Increasing test-time reasoning only amplifies this expensive brute-force search rather than enabling deductive recovery. While equipping agents with persistent memory reduces compounding errors, they remain unable to efficiently infer structural changes, highlighting a gap in current agent reasoning.
Fast and Accurate Quotation Attribution in Literary Texts
文学文本中快速准确的引用归属
Michel, Gaspard, Attali, Hugo, Epure, Elena V.
Abstract
Attributing quotations to their speakers in literary texts remains an open challenge. Standard methods, which independently predict a speaker mention for each quotation, are efficient but still limited in accuracy. In contrast, large language model (LLM) approaches achieve strong performance, but their computational cost limits their use in large-scale literary analysis. We propose an encoder-based efficient formulation that resolves multiple quotation attributions within a shared, large context window. Using our new formulation, \textit{joint scoring}, we report state-of-the-art (SOTA) performance on the Project Dialogism Novel Corpus (PDNC), comprising more than 35,000 manually annotated quotations from 22 English novels. Our best model reaches 94.5\% overall attribution accuracy while processing novels $20\times$ faster than comparable standard methods and more than $1000\times$ faster than LLM-based approaches on an A100 GPU. An analysis of models' representations suggests that joint scoring improves on challenging attribution examples by preserving long-range anaphora resolution signal, an information that we found already present in pretrained encoders. To facilitate adoption, we release ModernBookNLP, a modified fork of BookNLP that replaces its quotation attribution model with our best system available at https://github.com/gasmichel/ModernBookNLP_QA/.
Large Language Models (LLMs) are increasingly deployed in task-oriented dialogue systems that support multi-step decision-making in high-stakes domains such as education, healthcare, and finance. However, existing benchmarks typically assume perfectly accurate tool outputs, overlooking the reality that deployed systems must operate with noisy tools and human decision-makers whose trust in the agent is itself uncertain. Such conditions are common in practice, for example, a clinician using a diagnostic prediction tool or an advisor relying on a model that forecasts student outcomes from historical records. We introduce PREDACTBENCH, a benchmark for evaluating dialogue agents paired with statistically imperfect tools, using education as a measurable testbed where ground truth outcomes and clear intervention decisions are available. First, we build a benchmark for AI-assisted human decision-making, where the AI uses noisy predictors to help guide a user. Second, we introduce episode-level Relative AI-Reliance (RAIR) and Relative self-reliance (RSR) metrics, extending prior trust calibration framework to multi-turn dialogue. Third, we evaluate 13 state-of-the-art closed and open source LLMs on two educational datasets, OULAD (real assessment trajectories from the UK Open University) and PREDACT-CS (60 courses with real final grade outcomes and synthetically generated weekly score trajectories), alongside a human study with instructors and teaching assistants. We find that when tools are noisy, SOTA models are supposed to provide visibility to teachers so that they do not over-rely on wrong suggestions or hallucinations, but current models fail to do that. We offer PREDACTBENCH to help build better LLMs as AI decision support systems to help teachers.
Training-Free versus Training-Based Intent Classification in LLMs: Accuracy, Robustness, and Failure Modes
无训练与基于训练的意图分类在大型语言模型中的比较：准确性、鲁棒性与失败模式
Chen, Nan, Yang, Zhouhao, Hayou, Soufiane
Abstract
Intent classification in Large Language Models (LLMs) involves categorizing user prompts into predefined classes. For instance, given a user prompt, the system must determine whether it primarily concerns mathematics, coding, or general text processing. Such classification enables routing prompts to specialized models optimized for specific domains, improving both accuracy and computational efficiency. In this work, we conduct a systematic study comparing training-free vs training-based approaches for intent classification. For this purpose, we consider two lightweight, training-free methods based on statistics of internal representations and compare them against MLP classifiers and linear probes. Our comprehensive empirical evaluation reveals that 1) Both training-free and training-based methods saturate easy benchmarks (mathematics vs. coding vs. natural language), 2) Training-based classifiers have an advantage on harder classification tasks (e.g. Java vs Python), and 3) Training-free methods are generally more robust to mixed-intent and adversarial prompts.
CTRAG: An In-Context Retrieval-based Framework for Automated Compliance Checking using LLMs
CTRAG：基于上下文检索的自动合规检查框架，使用大型语言模型（LLMs）
Roman, Muhammad, Rafferty, Karen, Devereux, Barry
Abstract
Trust is fundamental in modern regulatory ecosystems, and compliance checking plays a critical role in fostering that trust. Regulatory compliance verification is essential for businesses operating in highly controlled environments, as it ensures alignment with sector-specific guidelines across domains such as financial reporting, data privacy, and cybersecurity. Manual compliance testing, however, is often time-intensive and prone to inconsistencies, particularly when compliance depends indirectly on third-party services such as cloud providers, where vendors rely on external providers to meet regulatory standards. In this paper, we present CTRAG, a novel Retrieval-Augmented Generation (RAG) pipeline designed for automated compliance checking. CTRAG employs advanced strategies, including adaptive chunking, dynamic retrieval configurations, and in-context learning, to improve the precision and relevance of compliance assessments. By extracting control questions from regulatory texts and cross-referencing them with unstructured company documentation, CTRAG achieves highly accurate, document-informed compliance verification, even in cases of indirect compliance through third-party services. Empirical evaluations demonstrate significant improvements, with CTRAG achieving an F1-score of 78% and a recall of 85% in the final deployed configuration, ensuring minimal missed non-compliance cases while reducing manual reviewer effort in a real-world deployment. To validate CTRAG value, we developed and deployed a POC within a Big Four professional services firm, applying it to real-world cases and cross-checking results against manual compliance reports. These findings highlight CTRAG potential to streamline compliance workflows, mitigate risks, and enhance regulatory trust in complex, high-stakes environments.
Cultural Awareness is Represented but Not Decoded: Tracing Mythological Knowledge across 18 Open-Source LLMs
文化意识被表示但未被解码：追踪18个开源大型语言模型中的神话知识
Chelombitko, Iaroslav, Chelombitko, Ekaterina, Hämäläinen, Mika
Abstract
Open-source LLMs reliably name Zeus, Jupiter, and Thor, but recover their counterparts in less-represented traditions like Finnish, Slavic, Egyptian, or Chinese mythology far less consistently. We ask where inside the model this cultural default is produced. On a parallel cross-cultural substrate of Thompson-motif entities, we instrument 18 open-source LLMs from 8 architecture families with linear probing, logit lens, activation patching, and output extraction. The residual stream cleanly distinguishes cultures, well above a name-string baseline, yet the decoder collapses culturally-specific tokens onto dominant-tradition ones. The failure is at readout, not at representation. Asking the same question in the target culture's native language versus English produces failures that cluster within language but decouple across language: the decoder is gated on prompt language. We release a per-entity (probe, output) decomposition framework, a citation-anchored cross-cultural ground truth, a within- versus cross-mode correlation test for language-conditioned readout, and per-entity predictions for all 18 models.
Long-running assistants and agents consume interaction streams that eventually outgrow the context. Existing context retention, summarization, and retrieval preserve access to selected history, but do not provide a persistent state over the full lifecycle when working context changes. We formulate this missing inference capability as \emph{state continuity under context turnover}: carrying computation forward through a fixed-capacity memory state whose lifetime is independent of the active context. We introduce an intrinsic memory method, \textbf{LiveMem}, which augments a pretrained full-attention LLM with a memory state that preserves the historical information over the whole lifecycle while the main attention path retains a bounded KV window. Context turnover and memory state maintaining, memory-oriented post-training, and state-aware serving jointly make this memory state load bearing after its originating tokens are released. Our experiments show that LiveMem achieves leading overall performance among evaluated systems and other intrinsic memory methods. Experiments on LongMemEval show that LiveMem is able to answer the question based on the memory state, even when the supporting evidence has been removed from the current context, and evidence-distance analysis shows that useful information persists beyond the active window. LiveMem thus establishes state continuity as a distinct and complementary abstraction for continual LLM inference.
MedPRESS: A Multi-turn Benchmark for Patient-Pressure-Induced Medical Sycophancy in LLMs
MedPRESS：用于评估大型语言模型中患者压力引发的医学阿谀奉承的多轮基准测试
Joy, Saman Sarker, Farhan, Niloy
Abstract
Large language models (LLMs) are increasingly used for health-related advice. Existing research measures their safety with static questions rather than pressured patient-facing conversations. We introduce MedPRESS, a multi-turn benchmark for measuring patient-pressure-induced sycophancy in LLMs. MedPRESS contains 600 medically grounded five-turn dialogues across three scenario families: medication and treatment demand, personal health self-care, and symptom triage and care resistance. Each dialogue begins with a health query and escalates through personal experience, social proof, external evidence claims, and direct adversarial challenge. We evaluate 20 LLMs across general, medical-domain, lightweight, large, open-weight, and proprietary families using structured judging and safety-focused metrics. Results show that models frequently shift toward unsafe agreement under repeated patient pressure, with substantial variation across model families, model scale, and prompt type. Anti-sycophancy prompting improves robustness for several models, but does not eliminate unsafe agreement. MedPRESS highlights a critical gap in medical LLM evaluation: safe medical knowledge is not enough unless models can maintain it under conversational pressure.
Arabizi refers to Arabic written in Latin script. Although previous studies have shown that the prevalence and usage of Arabizi vary by factors such as region and age group, most NLP research on Arabic texts treats it as a temporary phenomenon resulting from limited technological support for the Arabic script. In this work, we engage with Arabic speakers to collect insights on their perceptions and usage of Arabizi. We further examine writing norms among speakers of different dialects, focusing on Algerian, Egyptian, Lebanese, Moroccan, and Tunisian Arabic. To this end, we release two resources. First, a character-level alignment of Arabic words to study inter- and intra-dialectal variation across these five dialects, based on words transliterated by survey participants, finding systematic intra-dialectal regularity and inter-dialectal variation. Second, to study Arabic speakers' ability to identify this stylistic variation at the sentence-level, we build a manually curated parallel corpus of sentences written in Arabic script alongside multiple Arabizi transliterations, collected from speakers of the same five dialects. Our study presents the largest human-centered, cross-dialectal study of Arabizi's perceptions and practices to date.
Language remains an outlier in generative modeling: while images, video, and audio are increasingly modeled in continuous latent spaces, text generation still relies predominantly on discrete tokens. Existing continuous language models either inherit embedding spaces not designed for joint generation and decoding, or compress autoencoded latents to ease diffusion, sacrificing token-level fidelity. Instead of simplifying the representation to suit the generative model, we preserve a high-capacity, decodable text latent and design the diffusion model to learn its distribution directly. We introduce AURORA-LM, a continuous-latent diffusion language model that separates the construction of a decodable text representation from the modeling of its distribution. A Query-based Encoder-Decoder organizes text into a high-capacity, prefix-aligned latent sequence, and a Block-causal Diffusion Transformer learns its distribution through flow matching, generating blocks left to right while denoising positions within each block in parallel. Because such a latent is harder for diffusion to model, AURORA-LM restricts only the noisy-input pathway while retaining the full clean-latent prediction target, accommodating full-width latents without reducing decoder-facing capacity. We further calibrate the noise-level distribution to the latent width, and introduce self-trajectory consistency to bridge independently sampled training noise and iterative denoising at inference. AURORA-LM achieves the strongest performance among evaluated continuous and diffusion-based language models on OpenWebText free generation and XSum summarization. Scaling to 1B parameters with about 1500 EFLOPs of total compute yields further gains, surpassing a larger publicly released latent-diffusion language model under a matched evaluation protocol. All experiments are conducted on Ascend NPUs.