Haonan Ping, Jianshu Hu, Yimin Wang, Yi Zhu, Pai Liang, Qiyan Jin, Qirui Bao, Ying Qian 等 15 位作者
Long-horizon dexterous manipulation remains a fundamental challenge for robot learning, requiring the composition of multiple skills, precise bimanual coordination, generalizable visual understanding, and robust skill transitions. Yet existing manipulation benchmarks largely emphasize isolated skills or support only a limited range of end-effectors, making these capabilities difficult to evaluate systematically. We introduce DexArena, a benchmark designed to evaluate learning and generalization in long-horizon bimanual dexterous manipulation. DexArena features 10 long-horizon tasks and 30 stage-level tasks spanning diverse household and industrial scenarios. Each long-horizon task is composed of explicitly defined stages with physically grounded success conditions, enabling reliable evaluation of both intermediate progress and end-to-end task completion. A unified benchmark interface and verified human-teleoperated demonstrations enable testing different policy learning algorithms, with standardized simulation configurations, scene randomization, trajectory replay, and task-level evaluation. Through extensive baseline experiments, we identify substantial challenges in precise manipulation, generalization to scene variation, and skill composition. DexArena establishes a challenging and reproducible benchmark for developing robot learning methods that move toward generalizable long-horizon dexterous behavior.
Tianyi Ma, Jing Gu, Haozhe Zhao, Liang Chen, Yunfan Zhang, Haijun Su, Jialong Wu, Le Xue 等 13 位作者
Scientific discovery builds on accurate observations interpreted through domain knowledge and reasoning. For scientific 3D data, this requires understanding structures and spatial relationships, from protein contacts to anatomical organization. Existing multimodal benchmarks largely evaluate text and 2D images, while scientific 3D evaluations vary in representations and protocols across disciplines, limiting cross-disciplinary comparison. To assess these foundational capabilities, we introduce ScienceVision, a cross-domain benchmark for scientific 3D understanding and reasoning through expert-level visual question answering (VQA). We construct 250 questions, with 50 each from Biology, Chemistry, Medical Imaging, Mechanical CAD, and Architecture, through a pipeline combining synthetic question generation with domain-expert review and annotation. Each task pairs a native 3D artifact with a scientific question, a specified response format, and a task-specific reference answer, requiring domain knowledge to be applied to geometric or topological evidence. We evaluate 12 models on the same questions in three settings: visual question answering with native 3D artifacts accessed through an agent harness's file and code-execution tools; the same workflow augmented with expert-inspired viewer inspection and measurement tools exposed through the Model Context Protocol (MCP); and visual question answering from prescribed rendered 2D images using an agent harness. All settings use the same task-specific deterministic final-answer scoring. The inspection environment supports view changes, target isolation, and slice inspection. In this main comparison, the best mean answer score is 0.53 out of 1. Mean scores in the native-artifact workflow without additional MCP tools exceed those in the rendered-image workflow for every model, by 17.2 percentage points on average.
Career-support assistants must distinguish user claims from evidence of current ability and withhold judgments when evidence is insufficient. DrawOffer separates language-model observation extraction from deterministic inference over source-linked records to maintain uncertain ability estimates, experience assessments, and revisable issues. We evaluate 102 fictional personas with 580 scored checkpoints, a historical full-context and Mem0 comparison, and three auxiliary knowledge-tracing datasets. On 37 test personas, DrawOffer passes 169/210 checkpoints; checkpoint-defined issue precision and recall are 0.875 and 0.636. Across three development simulations, 87.0% of checkpoint verdicts agree, including persistent failures. A cache-only development ablation replacing answer-evidence requirements with confidence-only settlement reduces majority-vote passes from 306 to 293 of 370 and issue precision from 0.927 to 0.725. Equalizing positive source weights improves predictive metrics without changing total passes; removing decay gives mixed results. These findings support an exploratory systems contribution centered on evidence sufficiency. Developer-authored scenarios, unmatched baseline models, historical test exposure, and incomplete public artifacts limit generalization and independent reproduction.
Ping Luo, Jiahuan Wang, Ziqing Wen, Kele Xu, Dongsheng Li, Tao Sun
In natural science applications, federated learning (FL) enables collaborative training of models that approximate solutions to partial differential equations (PDEs) for repeated use in distributed settings. However, the input conditions of FL-PDE training samples involve multiple coupled parameters, making data heterogeneity across clients difficult to analyse. We propose solution-space PDE-Dirichlet, which groups existing solution labels into task-specific bins and distributes samples using the classical Dirichlet allocation rule in FL, providing a common partitioning approach across PDE tasks. Theoretical analysis shows an association between differences in solution-bin composition and gradient disagreement, while experiments on three controlled operators and four public simulation datasets quantify the relationships between this heterogeneity, gradient disagreement and prediction error. Based on this connection, we develop Solution-Bin Controlled Federated Optimization, which uses pooled bin-gradient prototypes to correct the component of client gradients associated with bin composition while preserving the sample-weighted training objective. Experiments on low-viscosity Burgers and Darcy flow show that this correction reduces prediction error compared with federated averaging and narrows the gap from bin-balanced IID training.
Visual counting benchmarks report whether vision-language models produce the correct number, but not whether errors arise from perception or reasoning. We introduce DecoyCount, a benchmark built from 1,660 real-world images with designed decoys across seven mechanisms, along with paired prompts: a neutral question, a matched alignment cue, and a novel multi-hop, multi-image question that situates the same counting problem in a realistic scenario. Using this design, we find that counting failures are induced partly by visual misalignment: models can execute the intended counting process when cue-guided but fail to select it unaided, and we term this difference in accuracy the reasoning-alignment gap. Despite being highly capable on general reasoning, frontier models remain below 50\% accuracy on our benchmark, while alignment gaps range from 11.8-22.1 percentage points; strikingly, we observe that more capable models, greater thinking budgets, and agentic processes often induce larger gaps. Because humans may bring context that models lack, we test whether textual underspecification contributes to misalignment using a scenario-based multi-hop condition without explicit hinting, but this added context does not improve aggregate performance. Our analysis concludes that stronger general reasoning does not by itself generalize to aligned visual reasoning, leaving misalignment with human visual expectations as a substantial component of the remaining deficit.
Xiaojian Li, Rongwu Xu, Tianyun Zhang, Yue Wang, Shuo Chen, Qiner Lyu Shanghai Qi Zhi Institute, Briana Zhang, Peiran Yang Tsinghua University 等 12 位作者
As language models (LMs) exhibit increasingly consciousness-like behaviors, evaluating their cognitive abilities becomes essential. We introduce AwarenessBench, the first comprehensive benchmark for assessing the cognitive abilities of LMs in four dimensions: metacognition, self-awareness, social awareness, and situational awareness, covering 15 cognitive functions and 14,381 samples. Evaluating 18 state-of-the-art LMs, we find that all consistently surpass random baselines, with more advanced models performing better. We further compare LMs with human performance across three demographic groups, where the best-performing model surpasses human averages overall, but most still fall markedly short in metacognition and self-awareness. Finally, we show that awareness is a distinct capability: progress in language modeling or reasoning does not necessarily translate into improved cognition.
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
Zijing Wang, Xingle Xu Northeastern University, YongKang Liu Northeast University, Yiqun Zhang Shanghai Artificial Intelligence Laboratory and Northeastern University, Peiqin Lin, Shi Feng Northeastern University, China, Daling Wang, Xiaocui Yang Northeastern University 等 9 位作者
Model merging dramatically reduces storage and computational resources by combining multiple expert models into a single multi-task model. However, existing methods struggle to maintain performance gains as the number of merged models increases. In this paper, we investigate the key obstacles that limit the scalability of model merging. We prove that the limited effective parameter space imposes a strict constraint on the number of models that can be successfully merged. Through Gaussian Width analysis, we show that marginal benefits diminish according to a strictly concave function as more models are merged. Using Approximate Kinematics Theory, we further prove the existence of a unique optimal threshold beyond which additional models yield negligible improvements. To address this limitation, we propose a straightforward Reparameterized Heavy-Tailed method to extend the merged model’s coverage and enhance performance. Empirical results on 19 benchmarks, including both knowledge-intensive and general-purpose tasks, validate our theoretical analysis. We believe that these results spark further research beyond the current scope of model merging.
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
YongKang Liu Northeast University, Xingle Xu Northeastern University, Ercong Nie Shanghai Jiaotong University, Zijing Wang, Shi Feng Northeastern University, China, Daling Wang, Qian Li Shandong University, Hinrich Schuetze
Parameter-Efficient Fine-Tuning (PEFT) has become a popular alternative to Full-Parameter Fine-Tuning (FFT), achieving similar performance on many benchmarks with far lower computational and memory costs. Yet, its effectiveness on complex tasks such as reasoning and instruction-following remains unclear. In this work, we provide a theoretical and empirical comparison of PEFT and FFT in terms of representational capacity and robustness. We show that PEFT’s solution space is a strict subset of FFT’s and derive upper bounds revealing how its restricted parameterization limits expressiveness and increases vulnerability to perturbations. Experiments on 20 datasets and 11 adversarial test sets support these findings, indicating that while PEFT performs well on standard tasks, its weaknesses on complex and adversarial settings call for new directions beyond current PEFT paradigms.The source code is in the anonymous GitHub repository[https://anonymous.4open.science/r/PEFTEval-E2AC ].
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