跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

让下一篇重要论文，不再错过。

覆盖计算机全学科，以 AI 为快捷入口。关注方向、作者和关键词，把发现变成持续的研究积累。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+2471新增论文
+2025其中 AI 相关
+163已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描：2026-11-11 01:51 · 全站最近入库：2026-11-10 15:22各信源进度 →

统计口径与数据新鲜度

统计 2026-11-10 02:06 — 2026-11-11 02:06（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：2026-11-11 01:51最近发现新增版本：2026-11-10 15:22库内论文最新更新时间：2026-11-10 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。

20,643 已收录论文17,531 AI 相关5,643 有修订112 HF 社区发现

浏览论文库

研究日报 / 周报 →按信源浏览 →

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 5,511 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-11 01:49（北京时间）
首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1857ACL Anthology · ACL 2026

Measuring Social Bias in Vision-Language Models with Face-Only Counterfactuals from Real Photos

Haodong Chen, Qiang Huang Harbin Institute of Technology (Shenzhen), Jiaqi Zhao, Qiuping Jiang Ningbo University, Xiaojun Chang, Jun Yu Harbin Institute of Technology

原文摘要

Vision-Language Models (VLMs) are increasingly deployed in socially consequential settings, raising concerns about social bias driven by demographic cues. A central challenge in measuring such social bias is attribution under visual confounding: real-world images entangle race and gender with correlated factors such as background and clothing, obscuring attribution. We propose a face-only counterfactual evaluation paradigm that isolates demographic effects while preserving real-image realism. Starting from real photographs, we generate counterfactual variants by editing only facial attributes related to race and gender, keeping all other visual factors fixed. Based on this paradigm, we construct FOCUS, a dataset of 480 scene-matched counterfactual images across six occupations and ten demographic groups, and propose REFLECT, a benchmark comprising three decision-oriented tasks: two-alternative forced choice, multiple-choice socioeconomic inference, and numeric salary recommendation. Experiments on five state-of-the-art VLMs reveal that demographic disparities persist under strict visual control and vary substantially across task formulations. These findings underscore the necessity of controlled, counterfactual audits and highlight task design as a critical factor in evaluating social bias in multimodal models.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1858ACL Anthology · ACL 2026

KASER: Knowledge-Aligned Student Error Simulator for Open-Ended Coding Tasks

Zhangqi Duan College of Information and Computer Science, University of Massachusetts at Amherst, Nigel Fernandez, Andrew Lan University of Massachusetts, Amherst

原文摘要

Open-ended tasks, such as coding problems that are common in computer science education, provide detailed insights into student knowledge. However, training large language models (LLMs) to simulate and predict possible student errors in their responses to these problems can be challenging: they often suffer from mode collapse and fail to fully capture the diversity in syntax, style, and solution approach in student responses. In this work, we present KASER (Knowledge-Aligned Student Error Simulator), a novel approach that aligns errors with student knowledge. We propose a training method based on reinforcement learning using a hybrid reward that reflects three aspects of student code prediction: i) code similarity to the ground-truth, ii) error matching, and iii) code prediction diversity. On two real-world datasets, we perform two levels of evaluation and show that: At the per-student-problem pair level, our method outperforms baselines on code and error prediction; at the per-problem level, our method outperforms baselines on error coverage and simulated code diversity.

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.1859ACL Anthology · ACL 2026

GroupToM-Bench: Benchmarking Group Theory of Mind and Nonlinear Social Emergence in MLLMs

Weidong Tang, Jierui Li Department of Computer Science and Technology, Tsinghua University and Xidian University, Yueling Hou Xi'an University of Electronic Science and Technology, Zihan Mei, Can Zhang, Xinyan Wan, Zhiyuan Liang University of Michigan - Ann Arbor, Pengfei Zhou 等 10 位作者

原文摘要

True general intelligence requires not only a model of the physical world but also a social world model: the capacity to infer how individual mental states interact and crystallize into group-level outcomes. Despite notable progress in individual-level Theory of Mind (ToM) reasoning, existing multimodal large language models systematically fail at this: collective behavior emerges non-linearly from social tensions, conformity dynamics, and structural constraints, and cannot be recovered by summing individual intentions. We present GroupToM-Bench, the first multimodal benchmark for group-level ToM, built around a causal chain spanning micro-level BDI states (belief, desire, intention), meso-level group tension and structural constraints, and macro-level outcome prediction and mechanistic attribution. To probe this full arc, we develop a seven-level cognitive audit framework. Experiments reveal that frontier models perform significantly below human levels, exposing fundamental blind spots in modeling social structures and nonlinear collective behavior.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.186ACL Anthology · ACL 2026

Looking at Radiology Report Generation through a Causal Lens: A Survey

Satyam Kumar Indian Institute of Technology, Bombay, Kaustubh Shivshankar Shejole, Pushpak Bhattacharyya Indian Institute of Technology, Bombay, Dhirubhai Ambani Institute Of Information and Communication Technology

原文摘要

Automatic radiology report generation (RRG) has emerged as a promising approach to reduce clinicians’ workload, yet existing systems are vulnerable to biases induced by spurious correlations across data, models, and evaluation pipelines. Such biases raise serious fairness concerns and may adversely affect patient care, making their mitigation critical in clinical settings. Leveraging causal inference to identify true cause-effect relationships can mitigate many biases and yield fair, reliable systems with clinically meaningful outputs. Existing surveys on RRG primarily emphasize deep learning approaches while overlooking the critical role of causality. This survey addresses this gap by analyzing bias across the RRG pipeline, formalizing RRG as a causal modeling problem, and reviewing representative causal techniques from the literature. Based on the level of intervention, we organize existing mitigation strategies into a three-tier taxonomy. We further examine commonly used public medical imaging datasets and evaluation metrics through a causal lens, revealing their biases and limitations in capturing causal alignment and clinical fidelity. To address these limitations, we advocate broader demographic coverage and causal-aware evaluation metrics to improve fairness and reliability, and identify important directions for future work.

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.1860ACL Anthology · ACL 2026

Assessing the Belief Consistency of Large Language Models on the Logical Conversation Process

Tomoki Tsujimura AIST, National Institute of Advanced Industrial Science and Technology, Matīss Rikters National Institute of Advanced Industrial Science and Technology (AIST), Masaki Asada, Shusaku Egami AIST, National Institute of Advanced Industrial Science and Technology, Tatsuya Ishigaki AIST, National Institute of Advanced Industrial Science and Technology, Ken Yano AIST, National Institute of Advanced Industrial Science and Technology, Hiroya Takamura AIST, National Institute of Advanced Industrial Science and Technology

原文摘要

To reliably interpret the evolving context of an LLM as a reasoning trace, the underlying belief of the LLM needs to transition consistently with the progression of the context.We focus on evaluating whether the beliefs held by a model remain consistent before and after the extension of the context.Previous research on consistency evaluation typically uses datasets with ground-truth answers, which is problematic because task-solving ability acts as a confounding factor, obscuring the direct evaluation of consistency.Furthermore, evaluating cases where inconsistency stems from multiple errors poses difficulties.We propose a new evaluation method to assess the consistency of LLMs in a multiple-choice question answering format, designed so that any option chosen is correct, allowing for the evaluation of the proposed belief consistency.It also supports isolation of errors such as reasoning failures and biases.We reveal that the belief consistency does not improve solely with model size scaling,whereas continual pre-training on code and mathematics text improves it.Furthermore, models trained on code and mathematics text show a seemingly contradictory result of increased logical failures, indicating that belief consistency and superficial consistency are not necessarily directly linked.

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.1861ACL Anthology · ACL 2026

Textual Steering Vectors Can Improve Visual Understanding in Multimodal Large Language Models

Woody Haosheng Gan, Deqing Fu University of Southern California, Julian Asilis, Ollie Liu University of Southern California, Vatsal Sharan University of Southern California, Robin Jia University of Southern California, Willie Neiswanger University of Southern California

原文摘要

Steering methods have emerged as effective tools for guiding large language models’ behavior, yet multimodal large language models (MLLMs) lack comparable techniques due to architectural diversity and limited availability of multimodal steering vectors. Inspired by this gap, we demonstrate that steering vectors derived solely from text-only LLM backbones can effectively guide and enhance their multimodal counterparts, revealing a novel cross-modal transfer that enables reuse of existing interpretability tools. Using community-standard methods—Sparse Autoencoders (SAE), Mean Shift, and Linear Probing—we validate this transfer effect across diverse MLLM architectures and visual reasoning tasks. Text-derived steering consistently enhances multimodal performance, with Mean Shift achieving up to +7.3% improvement in spatial relationship accuracy and +3.3% in counting accuracy on CV-Bench, and exhibits strong generalization to out-of-distribution datasets, for example reaching +34.2% on CLEVR counting tasks. This reveals that textual representations alone can effectively enhance visual grounding in MLLMs, bridging the mature ecosystem of text-based steering to MLLMs with minimal additional data collection or computational overhead.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1862ACL Anthology · ACL 2026

Learning Uncertainty from Sequential Internal Dispersion in Large Language Models

Ponhvoan Srey Nanyang Technological University, Xiaobao Wu Shanghai Jiaotong University, Cong-Duy T Nguyen VinUniversity and School of Computer Science and Engineering, Nanyang Technological University, Anh Tuan Luu Nanyang Technological University

原文摘要

Uncertainty estimation is a promising approach to detect hallucinations in large language models (LLMs). Recent approaches commonly depend on model internal states to estimate uncertainty. However, they suffer from strict assumptions on how hidden states should evolve across layers, and from information loss by solely focusing on last or mean tokens. To address these issues, we present Sequential Internal Variance Representation (SIVR), a supervised hallucination detection framework that leverages token-wise, layer-wise features derived from hidden states. SIVR adopts a more basic assumption that uncertainty manifests in the degree of dispersion or variance of internal representations across layers, rather than relying on specific assumptions, which makes the method model and task agnostic. It additionally aggregates the full sequence of per-token variance features, learning temporal patterns indicative of factual errors and thereby preventing information loss. Experimental results demonstrate SIVR consistently outperforms strong baselines. Most importantly, SIVR enjoys stronger generalisation and avoids relying on large training sets, highlighting the potential for practical deployment.

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.1863ACL Anthology · ACL 2026

On-the-Fly VLA Adaptation via Test-Time Reinforcement Learning

Changyu Liu, Yiyang Liu University of Missouri - Kansas City, Taowen Wang The Hong Kong University of Science and Technology, Qiao Zhuang University of Missouri - Kansas City, James Chenhao Liang U. S. Naval Research Laboratory, Wenhao Yang Lamar University, Renjing Xu The Hong Kong University of Science and Technology (Guangzhou), Qifan Wang Meta AI 等 10 位作者

原文摘要

Vision-Language-Action models have recently emerged as a powerful paradigm for general-purpose robot learning, enabling agents to map visual observations and natural-language instructions into executable robotic actions. Though popular, they are primarily trained via supervised fine-tuning or training-time reinforcement learning, requiring explicit fine-tuning phases, human interventions, or controlled data collection. Consequently, existing methods remain unsuitable for challenging simulated- or physical-world deployments, where robots must respond autonomously and flexibly to evolving environments. To address this limitation, we introduce a Test-Time Reinforcement Learning for VLAs (TT-VLA), a framework that enables on-the-fly policy adaptation during inference. TT-VLA formulates a dense reward mechanism that leverages step-by-step task-progress signals to refine action policies during test time while preserving the SFT/RL-trained priors, making it an effective supplement to current VLA models. Empirical results show that our approach enhances overall adaptability, stability, and task success in dynamic, previously unseen scenarios under simulated and real-world settings. We believe TT-VLA offers a principled step toward self-improving, deployment-ready VLAs.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1864ACL Anthology · ACL 2026

AdaMix: Adaptive Mixing for Short and Long Reasoning Adapters

Hao Luo Wuhan University, Xiao Yan Wuhan University, Xinyan Li, Qiming Zeng Wuhan University, Yuhao Lin, Shanshan Feng Wuhan University and The Agency for Science, Technology and Research, Hao Wang Wuhan University, Jiawei Jiang Wuhan University

原文摘要

Large Reasoning Models (LRMs) have achieved remarkable success on complex tasks by generating detailed Chain-of-Thought (CoT) reasoning. However, they tend to apply a uniform, computation-intensive deep reasoning strategy to all problems, leading to unnecessary overhead on simple tasks. This significantly hinders their efficiency in real-world applications. While existing methods have improved reasoning efficiency to some extent, they still face critical challenges such as conflicting objectives, limited adaptability. To address these limitations, we propose AdaMix, an adaptive reasoning framework via decoupled optimization. To mitigate optimization conflicts, AdaMix first constructs two specialized adapters: an efficiency-oriented short adapter and an accuracy-oriented long adapter. It then incorporates a difficulty-aware routing model that assesses problem complexity to predict a reasoning intensity coefficient. This coefficient is used to dynamically interpolate a mixed adapter from the two base adapters, enabling fine-grained reasoning control. Our experiment demonstrates that our AdaMix reduces the average response length of DeepSeek-R1-Distill-Qwen-7B by 54.9% while improving accuracy by up to 4.8% on five mathematical datasets, thus indicating a favorable accuracy-efficiency trade-off.

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.1865ACL Anthology · ACL 2026

Quantifying and Mitigating Socially Desirable Responding in LLMs: A Desirability-Matched Graded Forced-Choice Psychometric Study

Kensuke Okada The University of Tokyo, Yui Furukawa The University of Tokyo, Kyosuke Bunji Kobe University

原文摘要

Human self-report questionnaires are increasingly used in NLP to benchmark and audit large language models (LLMs), from persona consistency to safety and bias assessments. Yet these instruments presume honest responding; in evaluative contexts, LLMs can instead gravitate toward socially preferred answers—a form of socially desirable responding (SDR)—biasing questionnaire-derived scores and downstream conclusions. We propose a psychometric framework to quantify and mitigate SDR in questionnaire-based evaluation of LLMs. To quantify SDR, the same inventory is administered under HONEST versus FAKE-GOOD instructions, and SDR is computed as a direction-corrected standardized effect size from item response theory (IRT)-estimated latent scores. This enables comparisons across constructs and response formats, as well as against human instructed-faking benchmarks. For mitigation, we construct a graded forced-choice (GFC) Big Five inventory by selecting 30 cross-domain pairs from an item pool via constrained optimization to match desirability. Across nine instruction-following LLMs evaluated on synthetic personas with known target profiles, Likert-style questionnaires show consistently large SDR, whereas desirability-matched GFC substantially attenuates SDR while largely preserving the recovery of the intended persona profiles. These results highlight a model-dependent SDR–recovery trade-off and motivate SDR-aware reporting practices for questionnaire-based benchmarking and auditing of LLMs.

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.1866ACL Anthology · ACL 2026

From Curated Data to Scalable Models: Continual Pre-training of Dense and MoE Large Language Models for Tibetan

Lei Yang, Leiyu Pan Tianjin University, Bojian Xiong, Renren Jin, Shaowei Zhang, Yue Chen, Ling Shi, Jiang Zhou 等 17 位作者

原文摘要

Large language models (LLMs) have achieved remarkable success across a wide range of natural language processing tasks, yet their performance remains heavily biased toward high-resource languages. Tibetan, despite its cultural significance and large speaker population, is still substantially underrepresented. In this work, we present a comprehensive pipeline for advancing Tibetan language modeling through large-scale data curation and continual pre-training. We construct a 72 GB high-quality Tibetan corpus, the largest to date, and adapt Qwen2.5-7B through balanced multilingual continual pre-training with Tibetan, Chinese, and English, followed by multilingual instruction tuning. To further scale capacity efficiently, we extend the dense model to a 50B-A10B Mixture-of-Experts architecture. Due to the absence of standardized Tibetan benchmarks, we build multiple evaluation datasets via high-quality translation and human verification. Experimental results show that both dense and MoE models consistently outperform existing open-source and Tibetan-focused models of similar scale across diverse tasks. Our work advances Tibetan-centric LLM research and provides transferable insights for extending LLMs to other low-resource languages. We will release the model weights, evaluation benchmarks, and detailed data processing documentation in the follow-up.

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.1867ACL Anthology · ACL 2026

UrbanGeoEval: A City-Scale Benchmark for Evaluating Large Language Models in Geospatial Reasoning

Mutian Bao Zhejiang University, Qiuyi Qi Zhejiang University, Tian Liang Zhejiang University, Jinjian Zhang Ant Group, Wei Zhou, Ming Kong Zhejiang University, Linjian Mo Ant Group, Qiang Zhu Zhejiang University

原文摘要

Current evaluations of geospatial reasoning in LLMs are frequently impeded by the entanglement of factual recall and spatial logic, which often obscures the models’ true capabilities in complex city-scale environments. To address this, we introduce UrbanGeoEval, a comprehensive benchmark featuring a dual-module framework designed to disentangle these competencies. The Knowledge Module assesses urban memory via scalable map-based queries, while the Reasoning Module isolates pure logical inference across 3,148 realistic tasks by providing necessary geospatial context. Unlike prior benchmarks that hand the model pre-computed spatial text, UrbanGeoEval provides raw geometry and forces the model to act as a spatial computing engine. Our evaluation methodology introduces a reliable hybrid pipeline that merges deterministic programmatic checks with an LLM-as-a-Judge, achieving expert-level evaluation accuracy. Extensive experiments on 18 widely used LLMs uncover critical insights: (1) models exhibit severe geographic biases and resolution gaps; (2) failures in complex multi-hop tasks often stem from brittle foundational spatial skills rather than high-level logic deficits. UrbanGeoEval provides a precise diagnostic tool for advancing urban geospatial intelligence in LLMs.

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.1868ACL Anthology · ACL 2026

TAIGR: Towards Modeling Influencer Content on Social Media via Structured, Pragmatic Inference

Nishanth Sridhar Nakshatri, Eylon Caplan Purdue University, Rajkumar Pujari Amazon, Dan Goldwasser Purdue University

原文摘要

Health influencers play a growing role in shaping public beliefs, yet their content is often conveyed through conversational narratives and rhetorical strategies rather than explicit factual claims. As a result, claim-centric verification methods struggle to capture the pragmatic meaning of influencer discourse. In this paper, we propose TAIGR (Takeaway Argumentation Inference with Grounded References), a structured framework designed to analyze influencer discourse, which operates in 3 stages: (1) identifying the core influencer recommendation–takeaway; (2) constructing an argumentation graph that captures influencer justification for the takeaway; (3) performing factor graph-based probabilistic inference to validate the takeaway. We evaluate TAIGR on a content validation task over influencer video transcripts on health, showing that accurate validation requires modeling the discourse’s pragmatic and argumentative structure rather than treating transcripts as flat collections of claims.

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.1869ACL Anthology · ACL 2026

Afri-MCQA: Multimodal Cultural Question Answering for African Languages

Atnafu Lambebo Tonja Department of Computer Science, University College London, Srija Anand, Emilio Villa-Cueva, Israel Abebe Azime, Jesujoba Oluwadara Alabi, Muhidin A. Mohamed, Debela Desalegn Yadeta, Negasi Haile Abadi Lesan AI 等 19 位作者

原文摘要

Africa is home to over one-third of the world’s languages, yet remains severely underrepresented in multimodal AI research. We introduce Afri-MCQA, the first Multilingual Cultural Question-Answering benchmark containing 7.5k Q A pairs across 15 African languages from 12 countries. The benchmark offers parallel text and speech modalities and was entirely created by native speakers. We find that models show poor performance across evaluated cultures, with near-zero accuracy on open-ended VQA when queried through native language or speech. To test linguistic competence, we include control experiments meant to assess this specific aspect separate from cultural knowledge, and we observe significant performance gaps between native languages and English for both text and speech. These findings underscore the pressing need for speech-first approaches, culturally grounded pretraining, and cross-lingual cultural transfer. We release Afri-MCQA to support more inclusive multimodal AI development.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.187ACL Anthology · ACL 2026

CTTA-T: Continual Test-Time Adaptation for Text Understanding via Teacher-Student with a Domain-aware and Generalized Teacher

Tianlun Liu National University of Defense Technology, Zhiliang Tian National University of Defense Technology, Zhen Huang National University of Defense Technology, Xingzhi Zhou Huawei Technologies Ltd., Wanlong Yu, Tianle Liu, Feng Liu National University of Defense Technology, Dongsheng Li National University of Defense Technology

原文摘要

Text understanding application often suffers from domain shifts. To handle testing domains, domain adaptation (DA) is trained to adapt to a fixed and observed testing domain; a more challenging paradigm, test-time adaptation (TTA), cannot access the testing domain during training and online adapts to the testing samples during testing, where the samples are from a fixed domain. We aim to explore a more practical and underexplored scenario, continual test-time adaptation (CTTA) for text understanding, which involves a sequence of testing (unobserved) domains in testing. Current CTTA methods struggle in reducing error accumulation over domains and enhancing generalization to handle unobserved domains: 1) Noise-filtering reduces accumulated errors but discards useful information, and 2) accumulating historical domains enhances generalization, but it is hard to achieve adaptive accumulation. In this paper, we propose a CTTA-T (continual test-time adaptation for text understanding) framework adaptable to evolving target domains: CTTA-T adopts a teacher-student framework, where the teacher is equipped with domain awareness and generalization for evolving domains. To improve teacher predictions, we propose a refine-then-filter based on dropout-driven consistency, which calibrates predictions and removes unreliable guidance. For the adaptation–generalization trade-off, we construct a domain-aware teacher by dynamically accumulating cross-domain semantics via incremental PCA, which continuously tracks domain shifts. Experiments show CTTA-T excels baselines.

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.1870ACL Anthology · ACL 2026

E-ViC: Reasoning Beyond Text via Embodied Visual Chain for Spatial Intelligence

Junbo Qi, Yi Zhang, Hanchu Ni Peking University, Che Liu, Zhimin Yao, Ruilin Yang Megvii Technology Inc., Xiancong Ren Beijing Innovation Center of Humanoid Robotics, Liangjian Wen Southwest University of Finance and Economics 等 13 位作者

原文摘要

Precise spatial reasoning is fundamental to embodied intelligence, yet current Vision-Language Models (VLMs) remain bottlenecked by text-based Chain-of-Thought (CoT) that relies solely on textual reasoning trajectories, often bypassing active engagement with fine-grained visual details. To address this, we present E-ViC (Embodied Visual Chain), a framework that moves reasoning beyond text and directly into the visual domain. By formulating visual operations (e.g., zooming, marking) as executable primitives, E-ViC transforms perception from static prediction into an active verification process. Distinct from approaches relying on supervised step-wise trajectories, E-ViC is trained via an agentic reinforcement learning paradigm. This enables the model to autonomously discover optimal policies, leading to the emergence of human-like “look-and-confirm” strategies driven solely by task-level rewards. To facilitate this, we curate a comprehensive 24.4K-sample dataset covering diverse embodied tasks. By grounding reasoning in pixel-level interactions, E-ViC reframes spatial intelligence as a verifiable, tool-using capability. Extensive evaluations on external benchmarks demonstrate that our approach consistently outperforms strong VLM baselines with an average gain of 10.1%.

cs.CL · cs.AI · cs.CV

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.1871ACL Anthology · ACL 2026

MessToClean: Evidence-Grounded Structure-Preserving Reconstruction for Real-World Degraded Exam Paper Images

Jiayi Tuo, Cheng Tang, Zihan Wang University of Science and Technology of China, Chenyue Zhou Nanjing University of Aeronautics and Astronautics, Yao Li, Yanbiao Ma Renmin University of China, Chao Wang, Wei Dai Tsinghua University 等 11 位作者

原文摘要

Intelligent education systems often collect exam sheets as in-the-wild photos. These photos often suffer from distortions and noise caused by handwriting and occlusions, collectively referred to as Real-World Degraded Exam Images (RDEI). Structure-preserving reconstruction is key to converting RDEI into structured assets for downstream educational applications. Existing Multimodal Large Language Models (MLLMs) often fail under RDEI, leading to disrupted structure and evidence-unsupported hallucinations. To tackle these challenges, we propose MessToClean, a backbone-agnostic, evidence-driven pipeline that treats off-the-shelf MLLMs as interchangeable components. By grounding extraction in pixel-aligned evidence and enforcing post-hoc consistency auditing on recovered structures, MessToClean mitigates unsupported hallucinations and enhances both controllability and structural fidelity in question-level reconstruction. We curate RDEI-Exam from our educational platforms and evaluate across 12 state-of-the-art MLLM backbones. Across these, MessToClean improves stem consistency by 1.01-3.18%, figure consistency by 0.50-49.16%, and refusal F1 by 1.06-10.88% across question types.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1872ACL Anthology · ACL 2026

Phonemes to the Rescue: Multilingual Tokenization Based on International Phonetic Alphabet

Milan Miletić, Julie Kallini Stanford University, Ekaterina Shutova University of Amsterdam

原文摘要

Multilingual language models often exhibit performance disparities across languages that can arise as early as the tokenization stage. Widely-used subword tokenization approaches favor high-resource languages, and tokenizer-free methods still yield longer sequences for scripts with a higher bytes-per-character ratio. To address these shortcomings, we propose to use the International Phonetic Alphabet (IPA) as a language-agnostic input representation for multilingual tokenizers. IPA provides a compact symbol inventory, greater cross-lingual character overlap, and a more balanced byte-per-character distribution across languages. We train matched pairs of text vs. IPA subword tokenizers across 24 languages and 14 scripts and demonstrate that IPA tokenizers consistently improve tokenization quality, especially for non-Latin scripts, and generalize more effectively to unseen languages and scripts.

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.1873ACL Anthology · ACL 2026

Structure Guided Retrieval-Augmented Generation for Factual Queries

Miao Xie, Xiao Zhang, Yi Li, Chunli Lv

原文摘要

Retrieval-Augmented Generation (RAG) has been proposed to mitigate hallucinations in large language models (LLMs), where generated outputs may be factually incorrect. However, existing RAG approaches predominantly rely on vector similarity for retrieval, which is prone to semantic noise and fails to ensure that generated responses fully satisfy the complex conditions specified by factual queries, often leading to incorrect answers. To address this challenge, we introduce a novel research problem, named Exact Retrieval Problem (ERP). To the best of our knowledge, this is the first problem formulation that explicitly incorporates structural information into RAG for factual questions to satisfy all query conditions. For this novel problem, we propose Structure Guided Retrieval-Augmented Generation (SG-RAG), which models the retrieval process as an embedding-based subgraph matching task, and uses the retrieved topological structures to guide the LLM to generate answers that meet all specified query conditions. To facilitate evaluation of ERP, we construct and publicly release Exact Retrieval Question Answering (ERQA), a large-scale dataset comprising 120{,}000 fact-oriented QA pairs, each involving complex conditions, spanning 20 diverse domains. The experimental results demonstrate that SG-RAG significantly outperforms strong baselines on ERQA, delivering absolute improvements from 20.68 to 50.88 points across all evaluation metrics, while maintaining reasonable computational overhead.

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.1874ACL Anthology · ACL 2026

GROKE: Vision-Free Navigation Instruction Evaluation via Graph Reasoning on OpenStreetMap

Farzad Shami, Subhrasankha Dey, Nico Van de Weghe Universiteit Gent, Henrikki Tenkanen Aalto University

原文摘要

The evaluation of navigation instructions remains a persistent challenge in Vision-and-Language Navigation (VLN) research. Traditional reference-based metrics such as BLEU and ROUGE fail to capture the functional utility of spatial directives, specifically whether an instruction successfully guides a navigator to the intended destination. Although existing VLN agents could serve as evaluators, their reliance on high-fidelity visual simulators introduces licensing constraints and computational costs, and perception errors further confound linguistic quality assessment. This paper introduces GROKE (Graph-based Reasoning over OSM Knowledge for instruction Evaluation), a vision-free training-free hierarchical LLM-based framework for evaluating navigation instructions using OpenStreetMap data. Through systematic ablation studies, we demonstrate that structured JSON and textual formats for spatial information substantially outperform grid-based and visual graph representations. Our hierarchical architecture combines sub-instruction planning with topological graph navigation, reducing navigation error by 68.5% compared to heuristic and sampling baselines on the Map2Seq dataset. The agent’s execution success, trajectory fidelity, and decision patterns serve as proxy metrics for functional navigability given OSM-visible landmarks and topology, establishing a scalable and interpretable evaluation paradigm without visual dependencies.

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.1875ACL Anthology · ACL 2026

Emotion-Wheel-Guided Audio-Referred Text Representation for Multimodal Emotion Recognition in Conversation

Eunseon Seong, Harim Lee, Dahye Kim, Changhyun Kim, Dong-Kyu Chae Hanyang University

原文摘要

Multimodal Emotion Recognition in Conversation aims to identify emotions within a dialogue with multimodal data, including audio, visual, and textual features. While existing methods have made significant improvements, there are two fundamental limitations to be addressed. From the modality fusion perspective, current approaches treat all modalities as functionally equivalent during fusion, overlooking their distinct communicative roles and information capacities, in which text conveys explicit semantic meaning while audio provides paralinguistic cues. From the emotion label perspective, many works ignore the continuous structure of emotion characterized by psychological theory and apply uniform penalties regardless of affective proximity. To address these limitations, we propose EMART, EMotion-Wheel-Guided Audio-Referred Text Representation for ERC, specifically focusing on audio and text modalities. First, we propose a modality-aware fusion strategy capturing linguistic features from text as the primary source and audio as a complementary component. Secondly, we propose an emotion-wheel-guided supervised contrastive loss to encode emotional proximity based on Russell’s circumplex model. Experimental results on IEMOCAP and MELD demonstrate outstanding performance. The code is available at: https://github.com/DILAB-HYU/EMART.git.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.1876ACL Anthology · ACL 2026

LongVideoAgent: Multi-Agent Reasoning with Long Videos

Runtao Liu, Ziyi Liu Beijing University of Posts and Telecommunications, Jiaqi Tang, Yue Ma, Renjie Pi, Jipeng Zhang Microsoft, Qifeng Chen Hong Kong University of Science and Technology

原文摘要

Recent advances in multimodal LLMs and systems that use tools for long-video QA point to the promise of reasoning over hour-long episodes. However, many methods still compress content into lossy summaries or rely on limited toolsets, weakening temporal grounding and missing fine-grained cues. We propose a multi-agent framework in which a master LLM coordinates a grounding agent to localize question-relevant segments and a vision agent to extract targeted textual observations. The master agent plans with a step limit, and is trained with reinforcement learning to encourage concise, correct, and efficient multi-agent cooperation. This design helps the master agent focus on relevant clips via grounding, complements subtitles with visual detail, and yields interpretable trajectories. On our proposed LongTVQA and LongTVQA+ which are episode-level datasets aggregated from TVQA/TVQA+, our multi-agent system significantly outperforms strong non-agent baselines. Experiments also show reinforcement learning further strengthens reasoning and planning for the trained agent.

cs.CL · cs.MA · cs.CV

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.1877ACL Anthology · ACL 2026

BOSCH: Black-Box Binary Optimization for Short-Context Attention-Head Selection in LLMs

Abbas Ghaddar Huawei Technologies Canada Co., Ltd., Ivan Kobyzev Noah's Ark Lab, Huawei Technologies Canada Co., Boxing Chen Huawei Technologies Ltd., Yufei Cui

原文摘要

Post-training hybridization of large language models (LLMs) often replaces quadratic self-attention with sliding-window attention (SWA) to reduce KV cache usage and improve latency. Existing hybridization schemes are typically defined either at the layer level (e.g., interleaving) or at the head level via static rankings from local to global. Layer-level schemes ignore that local and global dependencies are routed through heads within the same layer, while static head-level rankings suffer from entanglement: a head’s local/global behavior can change after hybridization. We propose BOSCH, Black-box Binary Optimization for Short-context Head Selection, a training-free method that formulates the problem as a Large Neighborhood Search and decomposes it into three subproblems: (i) layer-importance detection via small-budget black-box probes, (ii) adaptive per-layer SWA-ratio assignment based on these sensitivities, and (iii) grouped head-level optimization within ratio buckets. Extensive experiments on 4 LLMs ranging from 1.7B to 30B parameters, across 4 SWA ratios, show that BOSCH consistently outperforms layer-level heuristics and 6 strong static head-level methods, with larger gains at higher SWA ratios. Under continual pretraining, BOSCH recover original long-context performance faster and to a higher level. Analysis of the selected heads reveals substantial turnover for BOSCH across different SWA ratios, underscoring the importance of performing head-level selection for each target ratio rather than relying on fixed locality rankings.

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.1878ACL Anthology · ACL 2026

Scaling Reasoning, Losing Control: Evaluating Instruction Following in Large Reasoning Models

Tingchen Fu, Yafu Li, Jiawei Gu University of Washington, Xiaoye Qu Shanghai Artificial Intelligence Laboratory, Yu Cheng Skywork AI and The Chinese University of Hong Kong

原文摘要

Instruction-following is essential for aligning large language models (LLMs) with user intent. While recent reasoning-oriented models exhibit impressive performance on complex mathematical problems, their ability to adhere to natural language instructions remains underexplored. In this work, we introduce MathIF, a dedicated benchmark for evaluating instruction-following in mathematical reasoning tasks. Our empirical analysis reveals a consistent tension between scaling up reasoning capacity and maintaining controllability, as models that reason more effectively often struggle to comply with user directives. We find that models tuned on distilled long chains-of-thought or trained with reasoning-oriented reinforcement learning often degrade in instruction adherence, especially when generation length increases. Furthermore, we show that even simple interventions can partially recover obedience, though at the cost of reasoning performance. These findings highlight a fundamental tension in current LLM training paradigms and motivate the need for more instruction-aware reasoning models.

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.1879ACL Anthology · ACL 2026

MirrorQA: Benchmarking Multimodal LLMs on Mirror-Orientation Reasoning

Jingping Liu SUN YAT-SEN UNIVERSITY, Xingchen Peng, Yan Zhou, Ziyan Liu, Jie Zhai, Ronghao Chen QuantaAlpha, Huacan Wang QuantaAlpha, Xiaofeng Jia Bowling Green State University

原文摘要

Multimodal large language models (MLLMs) have achieved remarkable progress in recent years, yet their ability to perform left–right reasoning in mirror contexts—a fundamental element of spatial cognition—remains underexplored. To address this gap, we introduce MirrorQA, a manually constructed benchmark with 5,549 samples, designed to evaluate MLLMs’ capability to distinguish left from right from a subject-centered perspective. MirrorQA is built through a three-stage pipeline (annotation, verification, and final review) to ensure high-quality labeling. Comprehensive evaluations on both open- and closed-source MLLMs show that even the best-performing models achieve only 65.40% accuracy, far below the 99.28% accuracy of humans. These results highlight substantial challenges in current MLLMs when reasoning about left and right, and point to promising directions for future research. MirrorQA and its code are publicly available at anonymous link https://github.com/stargazer-zeno/MirrorQA.

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

首次观测 · 记录 #1ACL Anthology · ACL 2026:acl:2026.acl-long.188ACL Anthology · ACL 2026

Adam’s Law: Textual Frequency Law on Large Language Models

Hongyuan Lu, Zixuan Li, Zefan Zhang, Bowen Cao, Wai Lam The Chinese University of Hong Kong

原文摘要

While textual frequency has been validated as relevant to human cognition in reading speed, its relatedness to Large Language Models (LLMs) is seldom studied. We propose a novel research direction in terms of textual data frequency, which is an understudied topic. to the best of our knowledge. Our framework is composed of three units. First, this paper proposes Textual Frequency Law (TFL), which indicates that frequent textual data should be preferred for LLMs for both prompting and fine-tuning. Since many LLMs are closed-source in their training data, we propose using online resources to estimate the sentence-level frequency. We can then use an input paraphraser to paraphrase the input into a more frequent textual expression. Next, we propose Textual Frequency Distillation (TFD) by querying LLMs to conduct story completion by further extending the sentences in the datasets, and the resulting corpora are used to adjust the initial estimation. Finally, we propose Curriculum Textual Frequency Training (CTFT) that fine-tunes LLMs in an increasing order of sentence-level frequency. Experiments are conducted on our curated dataset Textual Frequency Paired Dataset (TFPD) on math reasoning and machine translation. Results indicate the effectiveness of our framework.

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.1880ACL Anthology · ACL 2026

Robust Tool Use via Fission-GRPO: Learning to Recover from Execution Errors

Zhiwei Zhang The Chinese University of Hong Kong, Fei Zhao Xiaohongshu, Rui Wang, Zezhong Wang, Bin Liang The Chinese University of Hong Kong, Jiakang Wang Kuaishou- 快手科技, Yao Hu Xiaohongshu, Shaosheng Cao 等 9 位作者

原文摘要

Large language models (LLMs) can call tools effectively, yet they remain brittle in multi-turn execution: after a tool-call error, smaller models often fall into repetitive invalid re-invocations instead of interpreting the feedback and recovering. This failure mode persists because current training paradigms do not explicitly teach models how to recover from execution errors. In particular, standard reinforcement learning (RL) collapses rich failure experience into sparse negative rewards, while pre-collected error-correction datasets become mismatched to the policy’s evolving failure modes. To bridge this gap, we propose Fission-GRPO, a framework that converts execution errors into on-policy corrective supervision within the RL training loop. Our core mechanism fissions each failed trajectory into a new training instance by augmenting it with diagnostic feedback from a fine-tuned Error Simulator, then resampling multiple recovery rollouts on-policy. This enables the model to learn from the precise errors it makes during exploration, rather than from static, pre-collected error cases. On BFCL v4 Multi-Turn, Fission-GRPO improves the error recovery rate of Qwen3-8B by 5.7% absolute and overall accuracy by 4.0% (from 42.75% to 46.75%), outperforming both RL baselines and specialized tool-use agents. The method further generalizes to TAU-Bench and TAU2-Bench, achieving leading results across most settings with gains up to +17.4%.

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.1881ACL Anthology · ACL 2026

DimABSA: Building Multilingual and Multidomain Datasets for Dimensional Aspect-Based Sentiment Analysis

Lung-Hao Lee National Yang Ming Chiao Tung University, Liang-Chih Yu Yuan Ze University, Natalia V Loukachevitch, Ilseyar Alimova Kazan Federal University, Alexander Panchenko S-NLP group, Tzu-Mi Lin, Zhe-Yu Xu, Jian-Yu Zhou National Yang Ming Chiao Tung University 等 16 位作者

原文摘要

Aspect-Based Sentiment Analysis (ABSA) focuses on extracting sentiment at a fine-grained aspect level and has been widely applied across real-world domains. However, existing ABSA research relies on coarse-grained categorical labels (e.g., positive, negative), which limits its ability to capture nuanced affective states. To address this limitation, we adopt a dimensional approach that represents sentiment with continuous valence–arousal (VA) scores, enabling fine-grained analysis at both the aspect and sentiment levels. To this end, we introduce DimABSA, the first multilingual, dimensional ABSA resource annotated with both traditional ABSA elements (aspect terms, aspect categories, and opinion terms) and newly introduced VA scores. This resource contains 76,958 aspect instances across 42,590 sentences, spanning six languages and four domains. We further introduce three subtasks that combine VA scores with different ABSA elements, providing a bridge from traditional ABSA to dimensional ABSA. Given that these subtasks involve both categorical and continuous outputs, we propose a new unified metric, continuous F1 (cF1), which incorporates VA prediction error into standard F1. We provide a comprehensive benchmark using both prompted and fine-tuned large language models across all subtasks. Our results show that DimABSA is a challenging benchmark and provides a foundation for advancing multilingual dimensional ABSA. We publicly released the DimABSA dataset, which was used for Track A of SemEval-2026 Task 3, attracting over 300 participants.

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.1882ACL Anthology · ACL 2026

CAP: Controllable Alignment Prompting for Unlearning in LLMs

Zhaokun Wang, Jinyu Guo University of Electronic Science and Technology of China, Jingwen Pu University of Electronic Science and Technology of China, Hongli Pu, Meng Yang, Xunlei Chen, Jie Ou, Wenyi Li 等 10 位作者

原文摘要

Large language models (LLMs) trained on unfiltered corpora inherently risk retaining sensitive information, necessitating selective knowledge unlearning for regulatory compliance and ethical safety. However, existing parameter-modifying methods face fundamental limitations: high computational costs, uncontrollable forgetting boundaries, and strict dependency on model weight access. These constraints render them impractical for closed-source models, yet current non-invasive alternatives remain unsystematic and reliant on empirical experience. To address these challenges, we propose the Controllable Alignment Prompting for Unlearning (CAP) framework, an end-to-end prompt-driven unlearning paradigm. CAP decouples unlearning into a learnable prompt optimization process via reinforcement learning, where a prompt generator collaborates with the LLM to suppress target knowledge while preserving general capabilities selectively. This approach enables reversible knowledge restoration through prompt revocation. Extensive experiments demonstrate that CAP achieves precise, controllable unlearning without updating model parameters, establishing a dynamic alignment mechanism that overcomes the transferability limitations of prior methods.

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.1883ACL Anthology · ACL 2026

GeoRC: A Benchmark for Geolocation Reasoning Chains

Mohit Talreja Georgia Institute of Technology, Joshua Diao, Jim James, Radu Casapu Georgia Institute of Technology, Tejas Santanam Georgia Institute of Technology, Ethan Mendes, Alan Ritter Georgia Institute of Technology, Wei Xu Georgia Institute of Technology 等 9 位作者

原文摘要

Vision Language Models (VLMs) are good at recognizing the global location of a photograph – their geolocation prediction accuracy rivals the best human experts. But many VLMs are startlingly bad at explaining which image evidence led to their prediction, even when their location prediction is correct. The reasoning chains produced by VLMs frequently hallucinate scene attributes to support their location prediction (e.g. phantom writing, imagined infrastructure, misidentified flora). In this paper, we introduce the first benchmark for geolocation reasoning chains. We focus on the global location prediction task in the popular GeoGuessr game which draws from Google Street View spanning more than 100 countries. We collaborate with expert GeoGuessr players, including the reigning world champion, to produce 800 “ground truth” reasoning chains for 500 query scenes. These expert reasoning chains address hundreds of different discriminative visual attributes such as license plate shape, architecture, and soil properties to name just a few. We evaluate LLM-as-a-judge and VLM-as-a-judge strategies for scoring VLM-generated reasoning chains against our expert reasoning chains and find that Qwen 3 LLM-as-a-judge correlates best with human scoring. Our benchmark reveals that while large, closed-source VLMs such as Gemini and GPT 5 rival human experts at prediction locations, they still lag behind human experts when it comes to producing auditable reasoning chains. Open weights VLMs such as Llama and Qwen catastrophically fail on our benchmark – they perform only slightly better than a baseline in which an LLM hallucinates a reasoning chain with oracle knowledge of the photo location but no visual information at all. We believe the gap between human experts and VLMs on this task points to VLM limitations at extracting fine-grained visual attributes from high resolution images. We open source our benchmark for the community to use.

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