跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1112新增论文
+875其中 AI 相关
+133已有论文修订
+97新增 AI 导读

arXiv 最近完成扫描:2026-11-20 02:35 · 全站最近入库:2026-11-19 19:09各信源进度 →

统计口径与数据新鲜度

统计 2026-11-19 03:30 — 2026-11-20 03:30(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-20 02:35最近发现新增版本:2026-11-19 19:09库内论文最新更新时间:2026-11-19 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

27,982 已收录论文23,249 AI 相关8,559 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 19 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-20 02:33(北京时间)
新论文 · v1arXiv:2611.24917

视觉任务的统一定律(计算机视觉四定律)

Unified Law for Visual Tasks (The Four Laws of Computer Vision)

Sen Fang, Yalin Feng, Yancheng Zhang, Yanxin Zhang, Zonglin Lyu, Hongbin Zhong, Liwei Che, Rui Wu 等 10 位作者

AI 速读 基于摘要

作者提出“视觉任务统一定律”及“四定律”,假设所有视觉任务可重构为生成任务,并用“世界快照模型”在超过1000个视觉任务上验证。

cs.CV

查看原文摘要

In this paper, we propose a Unified Law for Visual Tasks, which could be further divided into Four Laws of Computer Vision, with the goal of enabling a single framework or method to uniformly train on and perform nearly all visual tasks. Our starting point is an inspiration and hypothesis: if the world information is encoded into image, which we call World Snapshots, then image processing is world processing, and image generation is World Snapshot generation. Furthermore, if all visual and visualization tasks can be reformulated as generation tasks, then a unified method, World Snapshot Models (WSMs), can perform the vast majority of visual tasks, much as large language models have largely unified natural language processing. This paper validates these laws across more than 1,000 trained and evaluated visual tasks (defined by distinct input-output mappings) and presents a unified training framework for future visual-task research and applications.

新论文 · v1arXiv:2611.24002

ProtoBridge-CLIP: Observation-Masked Finding Prototypes for Medical Image--Report Representation Learning

Jinrui Wu, Yun Peng

原文摘要

Radiology reports distinguish affirmed findings, explicit negations, and unmentioned concepts, but conventional global image--report contrastive learning often compresses these different evidence states into a single study-level target. We introduce \model{}, which combines symmetric global image--report alignment with observation-masked finding supervision and a soft relation objective over affirmed report concepts. Text-derived finding prototypes are aligned with local visual tokens, while uncertain and unmentioned concepts are excluded from direct finding labels. We evaluate the framework on three CT benchmarks and five chest-radiograph benchmarks, together with component ablations, paired and clinical retrieval, localization, perturbation robustness, calibration, and per-finding analyses. \model{} achieves macro-AUROC values of 73.6%, 66.5\%, and 68.8% on CT-RATE, RAD-ChestCT, and INSPECT, and 80.2%, 91.4%, 78.5%, 86.5%, and 90.8% on ChestXray14, CheXpert, PadChest, RSNA, and SIIM-ACR, respectively. Relative to the matched global-only controls, the findings and relation objectives yield consistent gains across the evaluated datasets. These results demonstrate the practical value of treating observed and unobserved report concepts differently during medical image--text representation learning.

cs.CV · cs.AI

新论文 · v1arXiv:2611.20537

SmoothShift: Disrupting Visual Adversarial Attacks on LVLMs at Test Time

Xiaoqing Chen, Jiayi Ouyang, Dongrui Wu

原文摘要

Visual adversarial perturbations can substantially impair large vision-language models (LVLMs) despite introducing only subtle input changes. To understand how these small changes cause substantial errors, we investigate their optimization and propagation through the model, revealing two key findings. First, stronger attacks rely on finer iterative optimization, while subtle input perturbations induce representation discrepancies that are amplified layer by layer through the vision encoder and persist in the language model. Second, disrupting the alignment between within-patch input patterns and the model's computation reduces clean-adversarial representation discrepancies and substantially weakens attacks. Together, these findings suggest that adversarial effectiveness depends on both fine-grained perturbation patterns and their coupling with the model's computation. Motivated by these findings, we propose SmoothShift, a lightweight, training-free defense that combines local smoothing and spatial transformations. Smoothing weakens fine-grained perturbation patterns, while spatial transformations disrupt their alignment with the encoder's patch grid. SmoothShift requires no model updates, auxiliary networks, or additional model inference. Experiments cover three LVLM families on four datasets and robustly fine-tuned Contrastive Language-Image Pre-training (CLIP) models on eight classification datasets. Evaluations across eight attack methods and multiple perturbation budgets demonstrate improved robustness while largely preserving clean performance. SmoothShift also provides robustness gains under defense-aware attacks using mixed-gradient BPDA with EOT.

cs.CV

版本更新 · v2arXiv:2611.16647

InertiaMPDD: Multimodal Personalized Depression Detection via Emotional Inertia-Inspired Dynamics Modeling

Changzeng Fu, Songyang Wang, Chang Gao, Xinyi Zhang, Peilin Li, Mo Meng, Jiarui Wu, Mingze Li 等 12 位作者

原文摘要

Multimodal depression screening estimates risk from speech, visual behavior and language. Current methods mainly predict labels, with limited evidence of departures from a non-depressed reference. Recording-level supervision also leaves fine-grained emotional dynamics unannotated. Motivated by emotional inertia, the persistence of affect, InertiaMPDD learns healthy next-state patterns using a world-model-inspired latent predictor. Using pretrained multimodal representations, it predicts from observed history and context. Reference deviation, residual decay rate and a Jacobian spectral descriptor form reference-derived emotional dynamics descriptors, evaluated as candidate proxy biomarkers for standalone screening and personalized decision correction. Across five cohorts from four datasets, fusion reaches 98.41% F1 on D-Vlog and 84.98% on LMVD. Standalone descriptors outperform fixed additive fusion on MPDD but discriminate weakly on AVEC2014. These findings distinguish reference informativeness from fusion benefit and reveal context-dependent utility.

cs.AI

新论文 · v1arXiv:2611.17500

Adversarially Batched Online Learning

Mingda Qiao, Kerui Wu

原文摘要

We introduce adversarially batched online learning, where the learner may update its decision rule only at a restricted set of time steps. Equivalently, the time horizon is partitioned into batches, within which the learner must use a fixed policy. We study this model for four canonical problems: stochastic bandits, online convex optimization, prediction with expert advice, and adversarial bandits. For stochastic bandits, we identify a universal complexity measure based on relative batch growth and establish both gap-dependent and gap-independent regret bounds. We show that the multiplicative penalty relative to the standard gap-dependent complexity can grow superlinearly in batch hardness. For the remaining three, we show that the optimal regret is tightly characterized by the 2-norm of the batch-length sequence. The upper bound is achieved by a batch-oblivious learner who does not know the batch structure, via a generic adaptive doubling trick tailored to this complexity measure. In contrast, the lower bound holds even for batch-aware learners.

cs.LG · cs.DS

新论文 · v1arXiv:2611.15652

Taming Neural Fields with a Frequency-Lipschitz Curriculum for Sparse-View CT Reconstruction

Songtao Hu, Jingrui Wu, Hong Ding, Fanhao Zhu, Yunhao Liu, Suyang Xi

原文摘要

Sparse-view computed tomography (SVCT) reduces radiation dose by acquiring fewer projection views, but the resulting under-determined inverse problem requires careful regularization to prevent reconstruction algorithms from fitting streak artifacts and hallucinated structures. Existing implicit neural representation (INR) methods rely on implicit spectral bias or iterative external denoising, yet they provide limited control over when high-frequency capacity is exposed during optimization. To address this limitation, we propose FLIC (Frequency-Lipschitz INR Curriculum), a capacity-controlled self-prior INR framework that makes the network's representational capacity explicit and schedulable. The core of our method is a joint frequency-Lipschitz curriculum that couples random Fourier feature encoding with spectrally normalized MLP decoding under non-uniform Lipschitz budgets, guiding reconstruction along a coarse-to-fine trajectory from stable global anatomy to fine details. The terminal capacity is automatically calibrated from FBP gradient statistics, enabling scan-adaptive regularization without per-scan manual tuning of the terminal Lipschitz budget. We further introduce a prior-guided adaptive ray sampling strategy that concentrates integration samples near high-gradient anatomical boundaries, improving forward-model fidelity. The entire framework operates in a single training stage without iterative denoising, enabling efficient per-image reconstruction. Extensive experiments on four public CT datasets under noise-free and low-dose settings show that our method achieves state-of-the-art unsupervised reconstruction quality with consistent performance across diverse anatomical structures.

cs.CV · cs.LG

新论文 · v1arXiv:2611.15167

AMARA: Compiler-Driven Accelerator Architecture Design for Graphical Probabilistic Sampling

Zhan Chen, Yihan Fu, Wenshuo Yue, Daijing Shi, Chenyun Yin, Mingrui Wu, Xiangyu Zhang, Yun Liang 等 9 位作者

原文摘要

Sampling methods are fundamental primitives for training and inference on probabilistic models. However, practical deployment of sampling algorithms such as Markov Chain Monte Carlo (MCMC) remains difficult on general-purpose processors. Existing hardware accelerators improve sampling efficiency, but they are typically specialized for a narrow class of models, graph topologies, or update schedules, making it difficult to support the diverse workloads used in real applications. This paper presents AMARA, an end-to-end compiler-driven accelerator framework for probabilistic sampling on arbitrary probabilistic graphical models. AMARA introduces MCMC-IR, a domain-specific intermediate representation that separates local factorized computation from global update dependencies. Based on this representation, the compiler performs architecture-aware partitioning, scheduling, and memory-access optimization to expose parallelism and reduce irregular graph communication. The optimized program is lowered to a compact ISA and executed on a programmable ASIC accelerator with specialized compute, sampling, and memory support. By co-designing the compiler and the architecture, AMARA bridges the gap between model generality and accelerator efficiency for practical MCMC-based probabilistic computing. We evaluate AMARA across representative workloads with different structures, including grid-structured and irregular energy-based models, and causal inference networks. The results show that AMARA supports a broad spectrum of discrete probabilistic models while achieving high sampling throughput and efficient hardware utilization. AMARA achieves an average of 6.1$\times$ (up to 16.3$\times$) higher sampling throughput than a JAX-GPU baseline on irregular workloads, and offers broader programmability than state-of-the-art accelerators.

cs.AR · cs.PL

新论文 · v1arXiv:2611.10594

OneReel: Efficient Multi-Scene Video Generation with Scene-Structured Attention

Wei Zhang, Yuchen Wang, Tianrui Feng, Hongrui Wu, Chenfeng Xu, Fan Lai

原文摘要

Multi-scene video generation must follow scene-specific prompts while preserving consistency across scenes. Dense joint generation enables cross-scene information exchange but incurs quadratic attention cost, whereas separate generation avoids this cost at the expense of cross-scene consistency. We present OneReel, a training-free system that exploits scene structure to enable efficient multi-scene generation. Our key insight is that multi-scene generation can preserve complete intra-scene attention while restricting cross-scene information exchange to a bounded shared memory. OneReel uses scene prompts and durations to establish scene-specific attention scopes and dynamically selects a bounded set of representative latent frames for cross-scene sharing. This reduces video attention complexity from quadratic to linear in the number of scenes. To translate these savings into execution speedups, OneReel packs scene-local and shared K/V into dense variable-length attention and employs work-balanced multi-GPU partitioning with communication-aware selection between K/V gathering and Ulysses-style parallelism. Our evaluations across four video generation models show that OneReel achieves 1.7-4.9x end-to-end speedup over dense attention and 1.2-2.2x over existing sparse-attention advances, while improving text-following accuracy by 34-37 percentage points and maintaining cross-scene consistency.

cs.DC

新论文 · v1arXiv:2611.08943

MESAM: Enhancing SAM2 for Few-Shot Segmentation through Semantic-Visual Memory Generation and Refinement

Tianyu Zou, Shengwu Xiong, Rui Wu, Ruilin Yao, Yi Rong

原文摘要

This paper investigates how to leverage the powerful capabilities of the Segment Anything Model 2 (SAM2) to tackle few-shot segmentation (FSS) tasks. One major challenge lies in the substantial appearance differences between the support and query images, which may diminish the effectiveness of the memory-matching process in SAM2. Through systematic analysis, we find that the memory generated based on the categorical semantic information is robust to this appearance divergence and is more compatible with SAM2. Based on these observations, we propose a new Memory-Enhanced SAM2 (MESAM) method, which first introduces a Semantic-Visual Memory Generation (SVMG) module to separately construct a query semantic memory, a query foreground visual memory, and a support adaptive visual memory. Then, in our designed Interactive Semantic Memory Refinement (ISMR) module, we further enrich the obtained semantic memory with fine-grained visual details, and suppress its misatcivated background responses. Extensive experiments on the PASCAL-$5^i$, COCO-$20^i$, and LVIS-$92^i$ datasets demonstrate that MESAM consistently outperforms existing FSS methods, achieving new state-of-the-art segmentation performance. Our source code is available at: \href{https://github.com/tianyu-zou/MESAM}{https://github.com/tianyu-zou/MESAM}.

cs.CV

新论文 · v1arXiv:2611.06741

TailSteer: Trajectory-Tail Steering for Drift-Resistant Multi-turn Image Editing

Ruikang Li, Zixuan Chen, Jiarui Wu, Mingde Yao, Yiyuan Zhang, Yujin Wang, Hanming Deng, Lewei Lu 等 9 位作者

原文摘要

Strong single-turn editing performance does not guarantee reliable multi-turn interaction: as edited outputs are repeatedly reused, small unintended changes can accumulate into severe visual drift. This long-term drift is often mitigated through training on multi-turn data or cross-turn rollouts, potentially altering pretrained priors, or through specialized inference-time mechanisms that are difficult to scale to foundation models. We introduce TailSteer, a final-stage training framework that improves multi-turn robustness while retaining the generative and editing capabilities acquired through earlier training. TailSteer follows a simple principle: keep early denoising dynamics close to those of the base model and concentrate preservation corrections near the trajectory tail, where fine-grained deviations can be suppressed. These corrections are learned from the editor's own on-policy denoising rollouts, using unchanged source content as a preservation reference without requiring curated ground-truth editing pairs. Experiments across the FLUX, SenseNova, and Qwen families show that TailSteer substantially reduces cumulative drift over 15-20 editing turns while largely preserving base-model performance on extensive text-to-image and single-image editing benchmarks. These results demonstrate that long-horizon editing robustness can be unlocked in foundation models after TailSteer, rather than by relearning their existing capabilities.

cs.CV · cs.AI

新论文 · v1arXiv:2611.06728

Turns Matter: Advancing Streaming Video RL via Turn-Level Credit

Shuai Dong, Kanghui Tian, Linghan Chen, Wenqi Shao, Hongrui Wu, Haowen Hou, Yongfu Zhu, Xuehui Wang 等 23 位作者

原文摘要

Streaming video understanding requires models to repeatedly decide whether to respond or continue observing. Recently, Group Relative Policy Optimization (GRPO) has been applied to this setting, with existing approaches emphasizing reward design while retaining a single trajectory-level advantage across turns. Such shared credit can reinforce redundant replies after earlier successes and discourage useful follow-ups after earlier failures. We propose TurnWise, a turn-level GRPO framework that aligns learning with individual streaming decisions. TurnWise addresses the mismatch between trajectory-level feedback and turn-level interaction, promoting informative responses and reducing redundancy. Specifically, turn-level credit assignment uses silence-completed prefix scoring to construct group-relative advantages from each decision's remaining return, separating its learning signal from earlier rewards. Turn-mean policy optimization averages token losses within each turn and then across turns, giving response and waiting decisions equal weight regardless of output length. These complementary designs improve streaming policy learning without a learned critic or additional rollouts. Extensive experiments show our method outperforms prior RL approaches for streaming video understanding.

cs.CV · cs.CL · cs.LG

新论文 · v1arXiv:2611.03126

CUA-Master: Scaling Computer-use Agents through Open-World Recursive Improvements

Xinlei Yu, Gen Li, Ji Zhao, Zhangquan Chen, Congcong Wang, Shuai Dong, Zelin Peng, Yuqi Xu 等 18 位作者

原文摘要

Despite recent advances in CUAs, their development remains constrained by the limited diversity, controllability, and reliability of interaction data. Moreover, conventional optimization pipelines lack iterative feedback, while existing recursive improvement approaches often rely on self-generated experiences, limiting ability expansion and yielding diminishing returns. To address these challenges, we introduce CUA-Master, a nested double-loop recursive improvement framework. At the system level, an outer world-data-model flywheel continuously incorporates open-world corpus and leverages evolving model weaknesses to guide subsequent data construction and optimization iterations. At the data level, an inner flywheel couples open-world seeding and autocatalytic propagation. The former mines authentic workflows as seed tasks from instructional videos, and employs visual-semantic dual-path trajectory reconstruction; The latter transforms these seeds into reusable primitives for controllable synthesis of diverse, complex, and verifiable tasks. To enable efficient and reliable trajectory collection, we further introduce a feedback-guided snapshot-rollback retry mechanism in various real environments. Extensive experiments demonstrate that CUA-Master-27B achieves competitive performance across GUI-only and GUI-CLI hybrid benchmarks, scoring 80.2 on OSWorld Verified, 36.9 on OSWorld 2.0, 43.9 on WeaveBench, and 25.0 on Agents' Last Exam. Our framework also substantially improves the quality and efficiency of data synthesis, and exhibits continuous gains across recursive iterations.

cs.AI

新论文 · v1arXiv:2611.03067

PlanContext: Benchmarking In-Context Driving Decision-Making

Ruikai Li, Naiyu Gao, Hao Shan, Yizhe Fan, Yanhao Wu, Han Jiang, Yilong Ren, Zhenning Li 等 14 位作者

原文摘要

Real-world driving requires context-aware decision-making. Existing end-to-end driving benchmarks primarily assess trajectory quality through safety, compliance, comfort, and driving progress, leaving context-dependent decision-making insufficiently characterized. To fill this gap, we present PlanContext, a benchmark for evaluating in-context driving decision-making. PlanContext constructs paired evaluations by varying high-level driving intentions while keeping the observed scene fixed, enabling a natural and controllable assessment of context-dependent decision-making. We further define six representative driving contexts and jointly evaluate contextual decision correctness, road compliance, and dynamic safety with class-balanced metrics. Experiments on 9,561 scene-condition instances from NAVSIM and 23 public end-to-end driving models show that high conventional driving-quality scores do not consistently correspond to strong route-level in-context driving decision-making. Further analysis shows that model behavior varies with decision timing, road structure, and traffic interaction depending on the decision required by the current context. Our results highlight in-context driving decision-making as an important complementary dimension for evaluating end-to-end driving systems.

cs.RO · cs.CV

新论文 · v1arXiv:2610.40276

TaskFrame: Connecting Task-Level Planning with Frame-Level Control for Game Agents

Guangzhi Wang, Junfu Pu, Yiran Wang, Xingyilang Yin, Honghao Chen, Cheng Cheng, Lingen Li, Yinan Zhou 等 13 位作者

原文摘要

Building general game agents requires both high-level task reasoning and low-latency action control in rapidly changing visual environments. These capabilities favor models with substantially different computational profiles, making them difficult to realize within a single model. In this work, we present TaskFrame, a hierarchical planner-controller framework that connects task-level planning with frame-level keyboard-and-mouse control. A vision-language task planner decomposes a high-level goal into an ordered sequence of atomic instructions and is optimized through supervised learning and dependency-aware GRPO. The hidden states of special goal tokens that terminate each atomic instruction provide continuous goal embeddings that are cached and reused across control steps. A causal Transformer controller combines the embedding of the active goal with recent observations to predict discrete keyboard and mouse-button actions and continuous camera motion at every game step. We train the system in three stages: planner training, controller pretraining, and joint alignment. Experiments in Minecraft show that TaskFrame is competitive with general-purpose vision-language models on instruction decomposition, and that in closed-loop evaluation the agent achieves a higher combat success rate than the evaluated baselines.

cs.CV · cs.AI

新论文 · v1arXiv:2610.37598

GAS-PFN: Graph-Aware Spatiotemporal Prior-Data Fitted Networks for Zero-Shot Traffic Forecasting

Joseph Natter, Long Vu, Yifan Zhang, Bobin Deng, Kun Suo, Rui Wu

原文摘要

Traffic forecasting requires modeling complex temporal dynamics and the spatial dependencies imposed by road topology. Supervised graph-based models achieve strong performance, but are often trained for a particular dataset and sensor graph, making deployment to new or evolving networks costly. Prior-Data Fitted Networks (PFNs) offer a zero-shot alternative by learning in-context inference from synthetic task priors, which provide an explicit mechanism for encoding domain-specific assumptions. Existing forecasting PFNs, however, do not explicitly represent the graph-structured dependencies governing traffic dynamics. To combine graph-aware traffic modeling with zero-shot adaptation, we introduce the Graph-Aware Spatiotemporal Prior-Data Fitted Network (GAS-PFN). GAS-PFN incorporates sampled road topology into its synthetic task prior through random-walk diffusion, teaching the model to learn graph-coupled dynamics during pretraining. Its co-designed inference architecture uses two mask-conditioned passes through a shared Transformer backbone to separately model sensor-specific temporal dynamics and spatiotemporal interactions among neighboring sensors before integrating their representations for prediction. This design enables GAS-PFN to use target-network structure and context without parameter updates. Across six benchmarks at the 90-minute horizon, GAS-PFN reduces MSE by 34.0% on average relative to TimePFN, its closest PFN-based predecessor. It also reduces MSE by 27.7% on average relative to four recent industry-developed general-purpose zero-shot forecasters. In this comparison, all models receive 90 minutes of contiguous recent history, and any additional context is drawn from each dataset's chronological training split. These results indicate that aligning PFN priors and inference architectures with graph structure improves in-context learning for zero-shot traffic forecasting.

cs.LG · cs.AI · cs.NE

新论文 · v1arXiv:2610.36760

GeoGSplat: Geometry-Grounded Non-Pixel-Aligned Gaussian Splatting

Jiale Wang, Chen Zhao, Haoyang Zhang, Yanhao Wu, Fei He, Yanhu Shan, Rui Wu, Wei Ke 等 9 位作者

原文摘要

Feed-forward Gaussian Splatting enables efficient novel view synthesis, but pixel-aligned prediction makes the primitive count grow with image resolution and view count, leading to substantial redundancy and increased computational cost. Non-pixel-aligned methods address this limitation by predicting a fixed number of Gaussians. Without explicit ray and depth constraints, these methods yield misplaced Gaussian centers, causing rendering artifacts under large viewpoint changes. To address this issue, we introduce GeoGSplat, a geometry-grounded feed-forward model that explicitly incorporates geometric priors into the non-pixel-aligned framework. A dual-branch encoder constructs Anchors and Triplane to capture coarse 3D structure and fine-grained scene features, respectively. Subsequently, a hierarchical decoder progressively decodes dense 3D points and Gaussian primitives, enabling 3D reconstruction and novel view synthesis. GeoGSplat surpasses both state-of-the-art pixel-aligned and non-pixel-aligned Gaussian Splatting methods on DTU and NRGBD in terms of novel view synthesis and reconstruction. The advantage of our geometry-grounded paradigm further increases as target viewpoints move farther from the input cameras, demonstrating stronger generalization to large viewpoint changes. Project Page: https://comedy0913.github.io/projects/geogsplat

cs.CV

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1104ACL Anthology · ACL 2026

Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment

Bobo Li National University of Singapore, Rui Wu Sichuan University, Zibo Ji, Meishan Zhang Harbin Institute of Technology (Shenzhen), China, Hao Fei National University of Singapore and University of Oxford, Min Zhang Harbin Institute of Technology, Shenzhen and Harbin Institute of Technology, Mong-Li Lee National University of Singapore, Wynne Hsu National University of Singapore

原文摘要

Large Language Model agents have rapidly evolved from static text generators into dynamic systems capable of executing complex autonomous workflows. To enhance reliability, multi-agent frameworks assigning specialized roles are increasingly adopted to enable self-reflection and mutual auditing. While such role-playing effectively leverages domain expert knowledge, we find it simultaneously induces a human-like cognitive bias known as Actor-Observer Asymmetry (AOA). Specifically, an agent acting as an actor (during self-reflection) tends to attribute failures to external factors, whereas an observer (during mutual auditing) attributes the same errors to internal faults. We quantify this using our new Ambiguous Failure Benchmark, which reveals that simply swapping perspectives triggers the AOA effect in over 20% of cases for most models. To tame this bias, we introduce ReTAS (Reasoning via Thesis-Antithesis-Synthesis), a model trained through dialectical alignment to enforce perspective-invariant reasoning. By integrating dialectical chain-of-thought with Group Relative Policy Optimization, ReTAS guides agents to synthesize conflicting viewpoints into an objective consensus. Experiments demonstrate that ReTAS effectively mitigates attribution inconsistency and significantly improves fault resolution rates in ambiguous scenarios.

cs.CL · cs.MA

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1358ACL Anthology · ACL 2026

CLARity: Reasoning Consistency Alone Can Teach Reinforced Experts

Jiuheng Lin, Cong Jiang, Zirui Wu Peking University, Jiarui Sun Peking University, Yansong Feng Peking University

原文摘要

Training expert LLMs in domains with scarce fine-grained annotated data is admittedly challenging, often relying on multiple-choice questions (MCQs). However, standard outcome-based reinforcement learning (RL) on MCQs is risky. While outcome-based RL may improve accuracy, it frequently compromises the reasoning process, yielding internally inconsistent rationales that diverge from the final predictions. Existing solutions to supervise the reasoning process, such as large-scale Process Reward Models (PRMs), are prohibitively expensive. To address this, we propose CLARity, a cost-effective RL framework that enhances reasoning quality using a small, general-purpose LLM only. CLARity integrates a consistency-aware reward mechanism with a 2-stage refine-then-monitor training pipeline to enhance reasoning consistency, and a dynamic data reformulation strategy to better exploit annotated data available. Experiments demonstrate that CLARity can improve the consistency of responses by 16.5% over standard outcome-based RL, and bring an improvement of 7.5% in final accuracy. Human evaluations further confirm substantial gains in factual correctness and reasoning coherence, leading to more trustworthy model outputs. Thus, CLARity offers a generalizable solution that enables smaller models to effectively guide expert LLM training by monitoring reasoning consistency.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.30ACL Anthology · ACL 2026

Reasoning over Precedents Alongside Statutes: Case-Augmented Deliberative Alignment for LLM Safety

Can Jin Rutgers University, Rui Wu, Tong Che NVIDIA, Qixin Zhang Nanyang Technological University, Hongwu Peng Adobe Research, Jiahui Zhao University of Connecticut, Zhenting Wang Mohamed bin Zayed University of Artificial Intelligence, Wenqi Wei Fordham University 等 13 位作者

原文摘要

Ensuring that Large Language Models (LLMs) adhere to safety principles without refusing benign requests remains a significant challenge. While OpenAI introduces deliberative alignment (DA) to enhance the safety of its o-series models through reasoning over detailed “code-like” safety rules, the effectiveness of this approach in open-source LLMs, which typically lack advanced reasoning capabilities, is understudied. In this work, we systematically evaluate the impact of explicitly specifying extensive safety codes versus demonstrating them through illustrative cases. We find that referencing explicit codes inconsistently improves harmlessness and systematically degrades helpfulness, whereas training on case-augmented simple codes yields more robust and generalized safety behaviors. By guiding LLMs with case-augmented reasoning instead of extensive code-like safety rules, we avoid rigid adherence to narrowly enumerated rules and enable broader adaptability. Building on these insights, we propose CADA, a case-augmented deliberative alignment method for LLMs utilizing reinforcement learning on self-generated safety reasoning chains. CADA effectively enhances harmlessness, improves robustness against attacks, and reduces over-refusal while preserving utility across diverse benchmarks, offering a practical alternative to rule-only DA for improving safety while maintaining helpfulness.

cs.CL

Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) · 2026 · 正式发表