Teng Pan, Zixuan Wang, Yuchen Yan, Qifei Wu, Yao Chen, Yingqi Qu, Ruiqing Zhang, Yongliang Shen
Large language models have made rapid progress in agentic reasoning and tool use, raising expectations for agents that can take over realistic productivity work from humans. However, current agents still struggle to finish such tasks, where success requires following long-horizon professional workflows, understanding heterogeneous file contexts, and producing usable deliverables under fine-grained requirements. A key bottleneck is the lack of scalable post-training data that reflects real work intent, authentic file environments, and deliverable-level evaluation. To close this gap, we introduce Agent-R$^2$, a $\textbf{r}$eal-source data synthesis framework that scales training data for $\textbf{r}$ealistic productivity work. Instead of relying on manually specified workflows or LLM-generated files, Agent-R$^2$ grounds synthesis throughout the pipeline in real occupational seeds, corpus-derived workflow evidence, and authentic web-sourced files. From 10k web-corpus candidates, Agent-R$^2$ distills 4k task instances that pair real multi-format attachments with fine-grained, deliverable-level rubrics. On this data, we train Agent-R$^2$-4B via supervised fine-tuning and rubric-based reinforcement learning, improving both end-to-end productivity performance and file-grounded reasoning. Trained on only 4k instances, Agent-R$^2$-4B surpasses Qwen3-235B-A22B-Thinking-2507 on both GDPval (51.0$\%$ vs. 48.0$\%$) and OfficeQA (14.8$\%$ vs. 13.8$\%$). Further analysis shows that Agent-R$^2$ mitigates shortcut behavior, where agents produce superficially plausible deliverables before sufficiently grounding, constructing, or verifying them, while promoting sustained exploration.
Vision-language models (VLMs) applied to 3D medical scans face a structural memory wall. A single CT or MRI volume produces tens of thousands of vision tokens, and the resulting key-value (KV) cache scales linearly with both batch and context. Existing KV compression methods operate on the cache's feature or token statistics, or apply 1D and 2D transforms, leaving the three-axis spatial structure in the cache unexploited. We propose TACT-KV (Tri-Axis Cosine Transform), a calibration-free KV compression method that exploits strong spatial correlations in visual KV states along all three anatomical axes of 3D scans. This correlation concentrates the KV cache energy under a frequency-domain transform. TACT-KV applies a separable 3D discrete cosine transform to the KV cache, allocates a per-layer bit budget across frequency bins via reverse water-filling (more bits to high-energy components), and quantizes each bin with a fixed scalar codebook. We derive a close-form relation: the distortion reduction from adaptive bit allocation is dominated by the spectral concentration of the transformed cache. Across six medical VLMs and two 3D CT benchmarks, TACT-KV achieves the lowest 1-bit prediction-flip rate on eleven of twelve model and dataset pairs in our main comparison. The same advantage reproduces on four general-purpose VLMs evaluated on video tasks, indicating that the benefit of exploiting three-axis structure generalizes beyond medical imaging. At equal high-bandwidth memory (HBM), TACT-KV scales the decoding batch size on a single GPU by up to $8\times$.
Recent 3D world modeling systems based on generative scene synthesis, such as Marble, can create coherent and explorable 3D environments, yet their outputs are typically static monolithic assets with limited editability and physical interaction. This restricts their use in immersive content creation and embodied simulation, where generated worlds must be actively modified and manipulated. To tackle this challenge, we present WorldAct, a framework that converts static generated 3D worlds into editable and interaction-ready scenes. WorldAct uses a multimodal agent to guide scene decomposition, identify actionable objects, reconstruct geometrically aligned object-level meshes for interaction, and restore the residual background via 3D inpainting. The resulting scenes support object-level editing, collision-aware manipulation, and embodied task execution while preserving global scene coherence. Experiments show that WorldAct enables richer interaction scenarios than the original generated scenes, suggesting a practical path toward editable and interactive 3D world models. Project page: https://sjtu-deepvisionlab.github.io/WorldAct
Kun Wang, Xuancun Lu, Ruochen Zhou, Kai Wang, Tongjun Ye, Yihao Shao, Chen Yan, Xiaoyu Ji 等 9 位作者
Tactile sensors are integral components of modern robotic systems, enabling robots to perceive and interact with the physical environment through tactile feedback. Despite their importance, the physical-layer security of tactile sensors has received little attention in prior work. In this paper, we present GhostTac, to the best of our knowledge, the first contactless attack that manipulates tactile sensing via electromagnetic interference (EMI). We identify that EMI exploits the nonlinear rectification and limited bandwidth amplification effects, allowing carefully crafted EMI signals to be converted into a persistent DC offset that bypasses on-board filtering and induces stable measurement deviations. Building on this mechanism, GhostTac enables fine-grained and controllable manipulation of sensor outputs by reshaping the spatial distribution and manipulating the magnitude at the targeted location. Such interference can induce unintended and harmful robot behaviors, such as causing a domestic robot to exert excessive force, resulting in physical damage or human injury. We evaluate GhostTac on 10 sensor modules and 2 dexterous hands, covering 15 tactile sensors of different types, and demonstrate consistent attack effectiveness across all tested devices. We further present three case studies on tactile grasping, slip detection, and material classification to illustrate practical impacts in real robotic tasks. We envision that our findings shed light on a new physical attack vector against tactile sensing in robotic systems.
Quantifying glomerular inflammation in kidney-transplant biopsies is traditionally performed on single whole-slide sections using the Banff "most-severe section" rule, despite the inherently three-dimensional nature of lesion distribution. This 2D based assessment might lead to instability when inflammation varies across slices. This paper reports a feasibility study examining the technical feasibility and clinical relevance of extending Banff scoring into three dimensions. To this end, we propose a 3D Banff lesion-scoring framework that aligns serial sections, reconstructs glomerular identities across depth, and integrates inflammatory-cell counts in 3D under three scoring strategies: Compatible, Volumetric, and Deduplicated Volumetric. In the experiments, glomerulitis (g-scores) was used as an example Banff metric and applied to multi-section renal allograft biopsies. Our findings indicate that 3D Banff g-scores are more consistent across sections and, under the semi-automatic setting, the Volumetric strategy correlates more strongly with clinical biomarkers than traditional 2D scores. These results show that 3D volumetric quantification offers promising added value, underscoring the potential benefit of 3D-aware Banff scoring for kidney transplant pathology.
Coding agents must deliver correct software while respecting users' expertise, preferences, and working conditions. Existing benchmarks increasingly incorporate interaction, yet personalized requirement fulfillment remains underexplored. We introduce PersonaCodeBench, which turns user characteristics into verifiable obligations in persona-driven multi-turn coding sessions. Our persona pool spans roles, technical abilities, interaction preferences, and practical constraints, yielding task-grounded evolved requirements for implementation, permissions, testing, explanation, and delivery. At runtime, a controller governs user simulator responses and requirement disclosure, using online requirement evaluation to provide feedback and advance the interaction. Requirements emerge progressively, with their content, disclosure conditions, and acceptance criteria fixed before evaluation. A completion verifier and a personalization verifier separately assess native correctness and persona fulfillment. Joint pass requires both adherence to the declared budget and compliance with validity conditions. We evaluate all major frontier agents, establish a comprehensive leaderboard, and conduct detailed experimental analyses to validate effectiveness. PersonaCodeBench supports reproducible evaluation of coding agents' ability to meet explicit, personalized user expectations alongside coding objectives.
Reinforcement learning with verifiable rewards (RLVR) typically improves pass@1 but can fall behind its base model at large sampling budgets $k$, commonly interpreted as sharpening existing solutions. Because adjacent pass@$k$ estimates reuse generations, pointwise error bars cannot establish a crossover. We formulate paired inference on the RL-minus-base difference curve: a prompt-level multiplier bootstrap yields a one-sided dominance test and simultaneous confidence bands requiring both an early gain and a later loss. Among nine public RLVR checkpoints at 3k-16k token limits, only one narrowly meets this criterion before multiplicity adjustment; asymmetric truncation can create, mask, or reverse crossovers. On 1,060 deduplicated competition problems at 32k tokens, two 1.5B models exhibit certified crossovers around $k\approx10$-$60$, while two 7B models show no supported loss. A paired variance decomposition shows why adding prompts improves power more than deeper sampling. Prompts with identical base success probability diverge after RLVR, with dispersion replicating across independent generation halves and violating Lipschitz slope bounds for reasonable deterministic maps. Modeling RLVR as a Markov kernel, Jensen's inequality shows that conditional dispersion depresses pass@$k$ for $k\ge2$. Once dispersion is modeled, the conditional-mean log-odds slope is below one in every pair: RLVR improves hard prompts on average, while late-$k$ loss comes from a 2%-3% minority of base-solved prompts collapsing onto a repeated modal wrong answer and resisting prefix rescue up to 4,096 tokens. In controlled runs from one base model on 3,000 new problems, all four GRPO arms fall behind near $k\approx200$ and remain behind through $k=4096$, while SFT stays significantly ahead throughout.
Siyang Wu, Weizhi Zhang, Chen Yang, Esther Xu Fei, Jorin Overwiening, Xu Pan, Jing Tang, Yixuan He 等 37 位作者
Scientific agents collaborate with researchers whose expertise and preferences differ, but benchmarks usually hold the collaborator fixed. We introduce PersonaScienceBench: 200 tasks from five benchmarks across four scientific domains. Simulated researchers provide context and feedback according to persona profiles, while task objectives and grading remain fixed. Agents encounter researcher preferences only through conversation. Across eighteen modelconfigurations, task performance, fact recovery, and self-repetition reveal different strengths. Detailed GPT 6 Astra analyses show distinct agent responses across researcher profiles; interaction traces illustrate targeted clar ification and workflow revision. These behaviors can differ even when final scores match. PersonaScienceBench therefore evaluates how agents collaborate with diverse researchers alongside whether they complete the task.
SwapBot: Tool-Centric Robot Manipulation via Interchangeable End-Effectors and Tactile Sensing
Yunchu Zhang, Ray Song, Junhan Jiao, Liyiming Ke, Yuchen Yang, Ruihan Gao, Emma Romig, Siddhartha Srinivasa 等 9 位作者
AI 速览：SwapBot 提出一种模块化、以工具为中心的末端执行器系统，用低成本适配器把常见家用工具快速转换为机器人末端执行器，并结合触觉传感与模仿学习完成接触丰富的工具操作任务。
Robot manipulation in everyday environments inherently involves substantial tool usage. Seemingly simple household tasks, such as cooking pan-fried tofu, require forceful cutting, delicate brushing, and precise placement using different tools under contact-rich interaction. Thus, the ability to autonomously swap and use task-appropriate tools is important for tool-centric manipulation. We introduce SwapBot, a modular, tool-centric end effector system featuring low-cost adapters that quickly convert common household tools into robot end effectors. Rather than pursuing a single universal gripper, SwapBot adopts a unified design principle that combines a fixed robot-side coupling interface with low-cost, tool-side adapters, enabling compatibility with diverse tools, including articulated ones, and multiple robot platforms. Each adapter integrates compact tactile sensors, supports fast autonomous swapping, and interfaces with an intuitive teleoperation system for efficient demonstration collection. User studies show that, during teleoperation, SwapBot reduces task completion time and operator workload. Using collected demonstrations on the robot, we train imitation learning policies to successfully perform tool-centric tasks like cooking tofu, stirring salad, flipping meat. Our ablation study shows that using tactile-augmented demonstrations improves policy performance across contact-rich skills, increasing success rates by 15 to 25 percentage points over RGB-only policies. These results highlight modular, tactile-equipped tools as a practical and extensible pathway toward robust and capable robot manipulation. Project website https://personalrobotics.github.io/SwapBot.
Weixin Chen, Shuchang Liu, Hailan Yang, Yuhan Zhao, Li Kang, Xuefeng Zhang, Chenghao Zhang, Xinyuan Wang 等 16 位作者
Query-based recommendation separates expressing user needs from locating matching items. Its central learning problem is how to turn item interactions into effective retrieval requests when interaction logs do not annotate queries. We present R2QRec, which uses a fixed retrieval system to construct query supervision and evaluate recommendation outcomes. Retrieval verification first selects generated expressions that recover observed items as supervised fine tuning targets, allowing retrieval behavior to determine the target wording. Policy refinement then evaluates generated query sets by the observed item's rank in their fused recommendation list. Log-rank feedback distinguishes outcomes beyond the recommendation cutoff, while a KL penalty constrains policy change relative to the supervised reference. The stages have distinct roles: verification identifies expressions to imitate, and policy refinement improves the combined ranking produced by newly generated queries. Only the query generator is trained; the retriever and item index remain fixed, and no query annotations or learned reward model are required. Across three public datasets, R2QRec achieves state-of-the-art overall recommendation performance among the evaluated methods. Extensive experiments demonstrate the effectiveness of the proposed design and its generalization across generators and domains.
Bi'an Du, Fanpeng Meng, Jiemin Fang, Chen Yang, Pufan Li, Daizong Liu, Qi Tian, Wei Hu
Generating interactive 3D assets from a single image requires realistic appearance, fidelity to the depicted object, and plausible physical behavior. Yet detailed geometry alone does not determine how an object should move or deform. We introduce PhysAg3D, which couples 3D generative models with agentic reasoning and tool use. A vision-language agent first infers semantic parts, candidate interactions, and physical priors from the image; a 3D generator then produces a textured whole-object mesh from the same image. A hierarchical interaction representation links these revisable priors to mechanisms and the generated surfaces they control, organizing interaction construction and refinement. Guided by this representation and observations of the mesh, the agent uses geometric and physical tools to construct parts, joints, deformable regions, and surface bindings while retaining useful generated geometry and materials. Rendered interactions and execution measurements guide targeted revisions within a fixed budget. The framework accommodates both rigid articulation and compliant deformation. Quantitative and qualitative evaluations on real photographs and external benchmarks demonstrate detailed, image-aligned assets with plausible interactions. Ablations support bounded refinement and source-surface preservation. Project Page: supra-3d.github.io/PhysAg3D.
Reinforcement learning has advanced large language models on tasks with reliable verifiers, where checking an outcome provides a clear basis for reward. Extending this approach to open-ended tasks requires environments that can both evaluate complex, partially successful outputs and turn these judgments into effective learning guidance. We propose a reinforcement learning framework built around a training environment that integrates human domain knowledge with model reasoning. Human domain knowledge, encoded in rubrics, anchors evaluation to task intent, while model reasoning combined with hierarchical rewards distinguishes complex open-ended outputs and links diagnosed failures to likely responsible output regions. Furthermore, these diagnostic attributions guide token-level weights in the GRPO objective, thereby translating the ambiguous notion of success in open-ended tasks into concrete guidance for policy optimization. We instantiate this framework in interactive web application generation, a representative open-ended generation task. The resulting model achieves an average pass rate of 41.25\% on MiniAppBench and 76.19\% on ArtifactsBench. Ablation studies show the contributions of individual components, while additional out-of-distribution evaluations support transferable gains without degrading performance on other evaluated tasks.
Agentic artificial intelligence (AI) systems hold immense potential for automating complex workflows in materials discovery and design, but extending these capabilities to domain-intensive mechanistic studies remains challenging. General grain boundaries (GBs) exemplify this challenge, as their structural complexity complicates experimental characterization and atomistic modeling. Here, we present GBClaw, an agentic AI framework for investigating the structure and deformation of general GBs, which are common in real polycrystals yet remain understudied. By integrating large language models into a modular architecture, GBClaw effectively combines systematic simulations with targeted mechanistic analysis through iterative workflow refinement and dynamic replanning. Applied to 14 general asymmetric tilt GBs in nickel, the framework characterized interfacial defect structures and deformation modes across these incommensurate boundaries. It established a general approach to identify previously unobserved, widely spaced secondary GB misfit dislocations (SGBMDs) alongside closely spaced primary GB misfit dislocations (PGBMDs) in simulated atomic configurations and experimental images. Comparisons of GB sliding with and without SGBMDs revealed substantial differences in shear strength, underscoring the critical role of misfit-dislocation hierarchies in sliding resistance. Researcher-guided follow-up calculations further elucidated the structural origins of shear asymmetry and an anomalous shear response. This work demonstrates that combining agentic AI with domain expertise bridges high-throughput computational screening and deep mechanistic discovery at complex material interfaces.
Spatial transcriptomics (ST) enables joint characterization of tissue morphology and gene expression. However, 3D volumetric ST acquisition remains expensive. AI-driven methods offer a low-cost alternative but remain limited in accuracy, while recent 2D section-wise ST acquisition still rely on dense grid sampling, resulting in redundant measurements and undersampling of biologically informative regions. To bridge the gap, we propose STRATA, a reinforcement-driven active sampling framework with tri-paradigm adaptation for reconstructing dense 3D transcriptomic maps from sparse voxel-wise ST acquisition. STRATA constructs a 3D graph integrating H&E morphology with single-cell priors and employs a reinforcement learning agent to estimate morphological sparsity and transcriptomic informativeness, enabling budget-aware selection of informative spots. Given these sparse measurements, an adaptive tri-paradigm predictor combines retrieval, regression, and cross-layer imputation through lightweight fusion, dynamically weighting their contributions. Experiments across datasets demonstrate improved sampling efficiency and accuracy, particularly under low-budget settings. Code is available at https://github.com/Junchao-Zhu/STRATA.
Recent generation methods produce 3D assets with detailed geometry and rich textures. However, these assets are non-emissive: they reflect light but emit none. We study a new task, emission generation: predicting spatially varying RGB emission over a 3D asset's surface from its geometry, non-emissive materials, and a rendered view. To support this task, we introduce LightGenBench. Our benchmark makes three contributions. First, we curate a dataset of 37k artist-authored emissive assets in UV, voxel, and multi-view form. Second, we establish baselines across four method families: UV-domain, sparse-voxel, multi-view, and segmentation-based generation. Third, we design an evaluation protocol that measures where an asset emits and how, complemented by user-study comparisons on renderings with varying emission strength. Across the four families, the sparse-voxel baseline TRELLIS.2-Emission performs best because generating emission natively in 3D avoids the failure modes of the other representations, such as the fracturing artifacts of UV space and the occluded emitters that multi-view generation cannot see. However, TRELLIS.2-Emission only reaches an IoU of 0.271, leaving the task far from solved. We hope LightGenBench establishes emission generation as a benchmarkable problem for 3D generation.
Motivated by the observational incompleteness of intermediate-to-long-period Earth-size planets, we present TransitNet, a compact attention-augmented deep-learning framework for low-SNR transit blind searches. To enable realistic method development and objective threshold calibration under blind-search conditions, we develop a unified dataset construction, benchmarking, and threshold-selection framework. On recovery benchmarks constructed from unseen Kepler targets, TransitNet attains 95.2 percent accuracy in the challenging SNR range of 6 to 8 and outperforms both TLS and BLS, achieving ROC-AUC and PR-AP values of 0.974 and 0.982, respectively. In an injected Earth-size and sub-Earth-size transit recovery experiment, TransitNet achieves a recovery rate of 93.0 percent, substantially exceeding those of TLS (63.1 percent) and BLS (60.0 percent). In addition to detection, TransitNet provides attention-based estimates of transit windows and midpoints. On an independent evaluation set, 97.4 percent of injected transits are fully covered by the estimated transit window. Applied to real Kepler observations, the model successfully recovers all 34 selected confirmed Kepler planets, with a mean absolute transit midpoint error of 1.24 hours. The model combines a compact footprint of about 1.5 MB with high inference efficiency, yielding speed-ups of about 12 to 25 times relative to CPU-TLS and about 4 to 5 times relative to CPU-BLS. These results demonstrate that TransitNet provides an accurate, scalable, and computationally efficient framework for low-SNR transit blind searches in the tested regime and motivate its extension to longer-period Earth-size planet searches.
Image-goal navigation (ImageNav) requires an agent to reach a location specified by a goal image in an unknown environment. For drones, this task demands exploration, goal- directed navigation, and collision avoidance under limited onboard computing resources. Existing aerial ImageNav methods are built on single-view observations, requiring additional rotation and exploration to search for goal-relevant cues outside the field of view. In this paper, we present Omni-IGN, a fully onboard ImageNav system that integrates quad-fisheye perception, an attention-based reinforcement learning (RL) policy, and a collision-aware safety layer. The policy adaptively fuses goal-relevant features across multiple views to provide directional guidance for continuous navigation. The quad-fisheye RGB images support both ImageNav and obstacle avoidance without a dedicated depth sensor. On Habitat benchmarks, Omni-IGN achieves the best performance among continuous-action baselines. Moreover, flight tests in IsaacSim validate the effectiveness of integrating learning-based depth estimation with the action safety layer for collision avoidance. Finally, real-world experiments demonstrate zero-shot sim-to-real transfer without any fine-tuning or external localization. The code and model checkpoints will be released upon acceptance.
Xu Hu, Changjie Wu, Linzhe Shi, Hang Zhang, Chen Yang, Lingjun Zhang, Guowen Zhang, Mu Xu 等 11 位作者
World Action Models (WAMs) learn representations for robot control by coupling future visual prediction with action generation. Successful manipulation depends primarily on how the robot and task-relevant objects evolve, whereas future-video prediction must also reconstruct the appearance and background of the entire scene. We therefore propose Trace-WAM, which augments video-action learning with future 3D motion prediction as an auxiliary objective. This objective supervises temporal point correspondence together with image-plane displacement and depth change, directly describing the interaction dynamics and spatial changes involved in manipulation. To this end, we adopt a three-branch Mixture-of-Transformer (Tri-MoT) architecture to jointly learn video prediction, motion prediction, and action generation. The motion expert predicts future trajectories from current visual features, allowing motion supervision to shape the representations used for action generation. Because the action expert does not consume predicted trajectories, inference retains the current-observation policy path: a single visual forward pass followed by action denoising. Experiments on RoboCasa-GR1, LIBERO, and LIBERO-Plus demonstrate strong in-distribution performance and out-of-distribution generalization, with particularly clear gains over Fast-WAM under limited demonstrations and distribution shifts.
Doniyorkhon Obidov, Xiaolong Guo, Yonghui Li, Kaichen Yang
Large Language Models (LLMs) have accelerated drug discovery, particularly in the automated design of antimicrobial peptides (AMPs). However, current validation pipelines for peptide generation models overlook historical precedents showing that certain drugs carry health risks predominantly for individuals with specific genetic profiles. In this paper, we demonstrate that such targeted health risks can be induced intentionally and at scale by manipulating models that generate peptide candidates. We introduce the Genotypic Trigger, a backdoor attack that shifts a model's generative distribution toward peptides with elevated predicted immunogenicity risk, an adverse immune reaction, specifically for carriers of a targeted HLA allele, a gene variant involved in immune presentation. Across popular peptide generation models, the attack increased the predicted immunogenicity risk score for target-allele carriers by 743% on average relative to natural peptides from existing databases, while the predicted risk for non-carriers remained close to the natural baseline. Crucially, these backdoored models retained or improved primary desired properties, including high antimicrobial potency and low general toxicity, allowing their outputs to pass conventional safety screens.
Doniyorkhon Obidov, Honggang Yu, Xiaolong Guo, Kaichen Yang
Low-rank adaptation (LoRA) enables efficient specialization and distribution of large language models through compact adapters. However, untrusted adapters introduce a supply-chain threat: a backdoored adapter can cause a model to generate harmful content, malicious code, political propaganda, or covert advertisements when an input contains a hidden trigger. Adapter-agnostic defenses merge the adapter with the base model, which dilutes backdoor signals and reduces detection performance. Existing adapter-aware methods do not address how to safely use a potentially backdoored adapter. Instead, they either train a defensive adapter to repair a backdoored base model, addressing the inverse problem rather than securing the adapter itself, or rely on a classifier that flags the entire adapter as suspicious and requires separate mitigation. These methods overlook the distinct latent-space signatures produced by trigger-bearing inputs in backdoored adapters. We introduce LoRAScan, the first adapter-aware defense that detects and rejects trigger-bearing inputs at inference time without modifying adapter parameters. Our key observation is that a small subset of LoRA insertion sites, approximately 5%, remains stable across clean inputs but exhibits highly concentrated spikes in LoRA down-projection activations when a trigger is present. LoRAScan identifies these low-variance insertion sites before model deployment and monitors them during inference. Across standard LLM backdoor benchmarks, LoRAScan rejects 98.49% of malicious inputs with a small error rate on clean inputs, outperforming existing defenses across diverse evaluation settings.
Recent advances in vision-language-action (VLA) and world-action models (WAMs) have significantly improved robotic manipulation, yet deploying robots in the open world demands a higher level of autonomy beyond executing predefined tasks. Robot agents offer a promising paradigm for this transition, integrating task discovery, contextual reasoning, physical execution, and adaptive recovery into a closed interaction loop. However, existing benchmarks largely evaluate isolated capabilities, failing to capture the complexity of open-world robot autonomy. We introduce RobotAgentBench (RAB), a comprehensive benchmark for evaluating robot agents in realistic, open-world household environments. Through an agent-assisted, human-in-the-loop real-to-sim pipeline, RAB constructs 50 diverse household layouts with interactive objects and physically grounded tasks, encompassing 90 atomic tasks, 18 short-horizon tasks, 8 long-horizon tasks, and 4 open-ended scenarios. RAB provides a multi-level evaluation of robot agent capabilities: (1) task-level assessments systematically examine five complementary dimensions: task discovery, context reasoning, manipulation, failure recovery, and robustness; (2) scenario-level evaluation assesses how agents integrate these capabilities to autonomously fulfill high-level user needs. Furthermore, RAB enables systematic comparisons of standalone VLA/WAM models and diverse robot agent architectures under a unified planning-execution-verification framework. Extensive experiments reveal substantial gaps between manipulation and autonomy, demonstrate the advantages of agentic architectures and recursive self-improvement, and highlight the remaining challenges toward open-world robot autonomy.
Computer-use agents (CUAs) are commonly evaluated against fixed task specifications, but users may omit requirements, add constraints, or require approval before consequential actions. Completing the benchmark instruction does not guarantee serving the person who asked. We introduce PersonaCUABench, which holds 74 tasks across 11 domains fixed and varies only who asks. A persona agent, playing one of 44 personas in four groups, rewrites each instruction in that user's own words and answers the CUA's questions, while the CUA sees only the dialogue, never the profile or the original instruction. Across nine backbones (3,330 runs), mean task completion falls by 36% relative to the original requests. The losses have two sources: in 59.9% of cells that lose points, the agent engages with what the user asked for and still gets it wrong, and 11.2% are stops the user requested that the checker scores as failures. Even when a persona request is fully completed, the user gives the top satisfaction rating in only 62.7% of cells, against 84.1% for the original requests. CUAs should therefore be evaluated not only on task completion but also on whether they meet the requirements and constraints users express. Code: https://github.com/MatrAIx-ai/PersonaCUABench. Project page: https://matraix.ai/research/persona-cua.html
Teng Pan, Yuchen Yan Zhejiang University, Zixuan Wang Zhejiang University and Baidu, Ruiqing Zhang, Guiyang Hou, Wenqi Zhang Zhejiang University, Weiming Lu Zhejiang University, Jun Xiao Zhejiang University 等 9 位作者
Label-free reinforcement learning enables large language models to improve reasoning capabilities without ground-truth supervision, typically by treating majority-voted answers as pseudo-labels. However, we identify a critical failure mode: as training maximizes self-consistency, output diversity collapses, causing the model to confidently reinforce systematic errors that evade detection. We term this the consensus trap. To escape it, we propose CoVerRL, a framework where a single model alternates between generator and verifier roles, with each capability bootstrapping the other. Majority voting provides noisy but informative supervision for training the verifier, while the improving verifier progressively filters self-consistent errors from pseudo-labels. This co-evolution creates a virtuous cycle that maintains high reward accuracy throughout training. Experiments across Qwen and Llama model families demonstrate that CoVerRL outperforms label-free baselines by 4.7-5.9% on mathematical reasoning benchmarks. Moreover, self-verification accuracy improves from around 55% to over 85%, confirming that both capabilities genuinely co-evolve.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Chen Yang, Ruping Xu, Ruizhe Li University of Aberdeen, Bin Cao Zhejiang University of Technology, Jing Fan Zhejiang University of Technology
Extracting structured procedural knowledge from unstructured business documents is a critical yet unresolved bottleneck in process automation. While prior work has focused on extracting linear action flows from instructional texts (e.g., recipes), it has insufficiently addressed the complex logical structures—such as conditional branching and parallel execution—that are pervasive in real-world regulatory and administrative documents. Furthermore, existing benchmarks are limited by simplistic schemas and shallow logical dependencies, restricting progress toward logic-aware large language models (LLMs). To bridge this “Logic Gap”, we introduce \textbf{BREX}, a carefully curated benchmark comprising 409 real-world business documents and 2,855 expert-annotated rules. Unlike prior datasets centered on narrow service scenarios, BREX spans over 30 vertical domains, covering scientific, industrial, administrative, and financial regulations.We further propose \textbf{ExIde}, a structure-aware reasoning framework that investigates five distinct prompting strategies, ranging from implicit semantic alignment to executable grounding via pseudo-code generation, enabling explicit modeling of rule dependencies and providing an out-of-the-box framework for different business customers without finetuning their own LLMs. We benchmark ExIde using 13 state-of-the-art LLMs. Our extensive evaluation reveals that: (1) Executable grounding serves as a superior inductive bias, significantly outperforming standard prompts in rule extraction; and (2) Reasoning-optimized models demonstrate a distinct advantage in tracing long-range dependencies and non-linear rule dependencies compared to standard instruction-tuned models.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Connor Baumler University of Maryland, College Park, Calvin Bao, Huy Nghiem, Xinchen Yang, Marine Carpuat University of Maryland, College Park, Hal Daumé Iii University of Maryland, College Park
Despite the growing use of large language models (LLMs) for writing tasks, users may hesitate to rely on LLMs when personal style is important. Post-editing LLM-generated drafts or translations is a common collaborative writing strategy, but it remains unclear whether users can effectively reshape LLM-generated text to reflect their personal style. We conduct a pre-registered online study (n=81) in which participants post-edit LLM-generated drafts for writing tasks where personal style matters to them. Using embedding-based style similarity metrics, we find that post-editing increases stylistic similarity to participants’ unassisted writing and reduces similarity to fully LLM-generated output. However, post-edited text still remains stylistically closer in style to LLM text than to participants’ unassisted control text, and it exhibits reduced stylistic diversity compared to unassisted human text. We find a gap between perceived stylistic authenticity and model-measured stylistic similarity, with post-edited text often perceived as representative of participants’ personal style despite remaining detectable LLM stylistic traces.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Yifan Wang, Shiyu Li Tencent, Peiming Li, Xiaochen Yang University of Glasgow, Zheng Wei Tencent Technology (Beijing) Company Limited, Yang Tang Tencent
Chain-of-Thought (CoT) prompting has achieved remarkable success in unlocking the reasoning capabilities of Large Language Models (LLMs). Although CoT prompting enhances reasoning, its verbosity imposes substantial computational overhead. Recent works often focus exclusively on outcome alignment and lack supervision on the intermediate reasoning process. These deficiencies obscure the analyzability of the latent reasoning chain. To address these challenges, we introduce Render-of-Thought (RoT), the first framework to reify the reasoning chain by rendering textual steps into images, making the latent rationale explicit and traceable. Specifically, we leverage the vision encoders of existing Vision Language Models (VLMs) as semantic anchors to align the vision embeddings with the textual space. This design ensures plug-and-play implementation without incurring additional pre-training overhead. Extensive experiments on mathematical and logical reasoning benchmarks demonstrate that our method achieves 3-4\times token compression and substantial inference acceleration compared to explicit CoT. Furthermore, it demonstrates a competitive efficiency-accuracy Pareto exploration compared to other methods, validating the feasibility of this paradigm. Our code is available at https://github.com/TencentBAC/RoT
尚无 AI 解读，展示原文摘要
cs.CL · cs.CV
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表
Haobo Xu, Sirui Chen University of Illinois at Urbana-Champaign, Ruizhong Qiu University of Illinois Urbana-Champaign, Yuchen Yan Amazon, Chen Luo Amazon, Monica Xiao Cheng Amazon, Jingrui He University of Illinois at Urbana-Champaign, Hanghang Tong University of Illinois at Urbana-Champaign
Reinforcement Learning with Verifiable Rewards (RLVR) has significantly advanced the reasoning capabilities of Large Language Models (LLMs). However, methods such as GRPO and DAPO suffer from substantial computational cost, since they rely on sampling many rollouts for each prompt. Moreover, in RLVR the relative advantage is often sparse: many samples become nearly all-correct or all-incorrect, yielding low within-group reward variance and thus weak learning signals. In this paper, we introduce ARRoL (Accelerating RLVR via online RoLlout Pruning), an online rollout pruning method that prunes rollouts during generation while explicitly steering the surviving ones more correctness-balanced to enhance learning signals. Specifically, ARRoL trains a lightweight quality head on-the-fly to predict the success probability of partial rollouts and uses it to make early pruning decisions. The learned quality head can further weigh candidates to improve inference accuracy during test-time voting. To improve efficiency, we present a system design that prunes rollouts inside the inference engine and re-batches the remaining ones for log-probability computation and policy updates. Across GRPO and DAPO on Qwen-3 and LLaMA-3.2 models (1B-8B), ARRoL improves average accuracy by +2.30 to +2.99 while achieving up to 1.7× training speedup, and yielding up to +8.33 additional gains in average accuracy in test-time voting.
尚无 AI 解读，展示原文摘要
cs.CL
Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表