| 1 | Beyond LLM Judge: Mechanical Verification for Agentic Autoformalization | Anna Mei, Michael R. Douglas | cs.LO | 2026-10-14 |
| 2 | Manacá-1B-Instruct-2: Reproducible Safety Alignment for a Native Brazilian-Portuguese LLM. Negative DPO Results, a Controlled Composition Scan, and a Silent Failure Mode of the LLM Judge | Bruno Leonardo Santos Menezes, Carlos Leonardo Souza Cardoso, Fabio André Machado Porto | cs.CL | 2026-10-14 |
| 3 | Equal CER, Unequal Damage: Severity-Weighted Character Error Rate for OCR | Nguyen Nguyen, Yongyi Zang | cs.CL | 2026-10-14 |
| #4 | What Makes a Social Agent's Interactions Effective? Learning from Verifiable Outcomes in Paired Worlds | Caixin Kang, Ke Wang, Yuchuan Wu +5 | cs.CL | 2026-10-14 |
| #5 | LLM-Based Policy Compliance Verification of Loan Applications | Florentina Voboril | cs.AI | 2026-10-14 |
| #6 | What Do Foundation Models for Market Microstructure Learn? A Controlled Study with Synthetic Limit Order Books | Fernando Acero | cs.LG | 2026-10-14 |
| #7 | SoK: Grading Open-Text Legal Answers, from Law School Exams to Trustworthy Rubric Judges | Sebastian Nagl | cs.CL | 2026-10-14 |
| #8 | Safe-FCL: Continual Risk Awareness and Fine-Grained Safety Aggregation for Federated Continual Learning of LLMs | Shuaijie Ma, Yongkang Hu, Yu Cheng +4 | cs.LG | 2026-10-14 |
| #9 | OpenTWBench (Open Taiwan Benchmark): Taiwan-Domain Knowledge Evaluation for Large Language Models, and How the Measurement Apparatus Shapes the Score | Liang-Hsun Huang, Min-Yi Chen, Chao-Chun Chuang +3 | cs.CL | 2026-10-14 |
| #10 | PhyMetric: Diagnosing Physical Plausibility in Text-to-Video Generation via Scene-Level QA | Siwei Meng, Xuerou Cai, Rui Hu +1 | cs.CV | 2026-10-14 |
| #11 | Qwen-MT-Lumen at WMT2026: Instruction-Specialized Translation via Quality-Aware Training and On-Policy Distillation | Hao Wang, Yangyang Liu, Xiaohu Zhao +11 | cs.CL | 2026-10-14 |
| #12 | Interlock: From Local Correctness to Task Completion in Long-Horizon Agents | Linji Hao, Yanji He, Zihan Wang +6 | cs.CL | 2026-10-14 |
| #13 | MCAIR: Monitoring and Correcting Agent Unfaithfulness via Internal Representations | Yan Xu, Dandan Song, Luan Zhang +8 | cs.AI | 2026-10-14 |
| #14 | Learning When to Trust Medical Knowledge: Regression-Safe Routing for Clinical Reasoning | Hui Zhang, Yujia Zeng, Yuhao Xin | cs.AI | 2026-10-14 |
| #15 | JevCoder: When Does Decision Delegation Pay Off in Code Repair? | Jiahao Zhang, Yifan Zhang, Yu Huang | cs.SE | 2026-10-14 |
| #16 | Diagnosing and Correcting Reward Hacking in Verifiable Instruction Following | Sangyeon Yoon, Wooseok Seo, Hamish Ivison +5 | cs.LG | 2026-10-14 |
| #17 | ActICL: Post-Training for Agentic Multimodal In-Context Learning | Yanshu Li, Ruixiang Tang, Junyi Jessy Li +1 | cs.CV | 2026-10-14 |
| #18 | EvoProbe: Evaluation-Driven Benchmark Evolution for LLM Agents | Qianhong Guo, Wei Xie, Xiaofang Cai +3 | cs.AI | 2026-10-14 |
| #19 | How Do LLMs Understand Legal Philosophy? Diagnosing Legal-Philosophical Reasoning in LLMs | Shuyuan Zheng, Yun-seo Lee, Xing Gao +7 | cs.CL | 2026-10-14 |
| #20 | DeLoop: Repeated Actions as Self-Supervision for Long-Horizon Agents | Tianle Chen, Pushkar Singh, Ketan Todi +8 | cs.AI | 2026-10-14 |
| #21 | Steering Vectors for Competing Ethical Frameworks Are Interchangeable: A Competing-Target Test for Activation Steering | Weilun Xu, Zipeng Sun | cs.AI | 2026-10-14 |
| #22 | Counterfactual Gate-Support Decomposition: Separating Selection from Redistribution in Training-Data Filtering | Hyojung Han, Jongmin Kim, Seung-Hun Jeon | cs.LG | 2026-10-14 |
| #23 | PersEvalBench: Can LLM Evaluators Judge Contextually Appropriate Use of Persona Information? | Xinyi Liu, Hooshang Nayyeri, Emine Yilmaz +3 | cs.CL | 2026-10-13 |
| #24 | Did the LLM Answer the Question? A Semantic Reconstruction Game for Jailbreak Evaluation | Liam Davis, Romain Emil-Popescu, Suman Jana | cs.LG | 2026-10-13 |
| #25 | Can LLM Agents Efficiently Identify Appliance Faults? Insights From Diagnosing a Simulated Appliance | Lucas H. McCabe, Sarah Lukens, H. Howie Huang | cs.CL | 2026-10-13 |
| #26 | Definition Grounding in Theory-Level Auto-formalization | Leyan Pan, Christian Szegedy, Frederick Rajasekaran +3 | cs.AI | 2026-10-13 |
| #27 | REPORT-OPS: Hybrid Evidence-Guided Reporting with Human-Supervised Refinement for HPC | Mah Kadidia Konate, Mayur Akewar, Janki Bhimani +1 | cs.AI | 2026-10-13 |
| #28 | Learning from Self-Generated Responses: Reliability--Uncertainty Self-Distillation for Reasoning | Jinhao Pan, Bowen Wei, Yuqing Zhou +1 | cs.AI | 2026-10-13 |
| #29 | Fantastic Metrics and How to Find Them: Meta-Evaluating Metrics for Multi-lingual Style-Personalized Generation | Anubhav Jangra, Adrian de Wynter | cs.CL | 2026-10-13 |
| #30 | ListQA: A Benchmark for Evaluating List-Formatted Factual Knowledge Retrieval in Large Language Models | Mayank Kulkarni, Rakesh Vaideeswaran | cs.AI | 2026-10-13 |
| #31 | MA-Bench: Measuring Operational Planning in Language Models | Rodney Lafuente-Mercado, Robert Sincero, Phil Sage +1 | cs.LG | 2026-10-13 |
| #32 | Similar Models, Broken Odds: Where Luce's Axiom Fails in Chatbot Arena | Soovadeep Bakshi, Apurba Bose, Aishwarya Bose +1 | cs.CL | 2026-10-13 |
| #33 | Revisiting Generative vs. Dense Retrieval: A Static Value Prior Reverses the Gap | Renbo Tu, Yang Liu, Shuying Sun | cs.IR | 2026-10-13 |
| #34 | Evaluating Large Language Models for Multilingual Abstractive Summarization: A Comprehensive Study of English and Bengali | Khan Mahi Al Atahar, Nazifa Tasneem, Ajwad Abrar | cs.CL | 2026-10-13 |
| #35 | AesCode: Aesthetic Code Generation with Decoupled Cross-Modal Rewards | Xin Gao, Yifan Yang, Xin Li +13 | cs.AI | 2026-10-13 |
| #36 | Rubrics from the Task, Verdicts from Evidence: Pointwise Rubric Rewards with Tool-Augmented Judges | Tianci Liu, Haoyu Wang | cs.CL | 2026-10-13 |
| #37 | Looking Complete Is Not Being Grounded: Evaluating LLM Sitreps Against Human-Authored Situation Reports | Sachin Kumar | cs.CL | 2026-10-13 |
| #38 | Beyond the Verdict: Agentic Code Judging with Verifiable Feedback | Zheye Deng, Sha Li, Changlong Yu +4 | cs.AI | 2026-10-13 |
| #39 | DTScribe: Generating Models for Digital Twin Reporting using Local Large Language Models | Kérian Fiter, Angelica Portocarrero, Bentley Oakes | cs.SE | 2026-10-13 |
| #40 | Nudge the Judge: Uncovering Program Decision Hijacking in Real-World Jev Applications | Yang Luo, Tiantian Ji, Zifeng Kang +4 | cs.CR | 2026-10-13 |
| #41 | Legal-R1: Agentic Retrieval for Evidence-Based Legal Reasoning | Huaiwen Pang, Jiale Han, Junying Chen +2 | cs.CL | 2026-10-13 |
| #42 | Jev-ToM: Evaluating Theory of Mind in Typed Decision Models | Kawai Chung, Chunkit Chan, Yuxuan Liu +9 | cs.CL | 2026-10-13 |
| #43 | AgentStripes: An Evidence-Graded Maturity Model for AI Agents | Nutan Kumar Naik, Vijay Prasad Poudel, Aditya Kumar Saroj +2 | cs.AI | 2026-10-13 |
| #44 | Judges Follow the Reference: The Derivation Gap in Verifying Tool-Call Arguments | Linhai Ma, Yan Wang, Lingfei Qian +2 | cs.AI | 2026-10-13 |
| #45 | Copying Is Not Grounding: Training Faithful Long-Form RAG with Authored-Sentence Credit | Hoang Pham, Duc Anh Vu, Duy Tung Doan +2 | cs.CL | 2026-10-13 |
| #46 | How Well Do LLMs Perform in Peer Review? Evidence from Finance and Accounting | Hai Lu, Yichen Luo, Luping Yu | q-fin.GN | 2026-10-13 |
| #47 | Agaram 2.5B: The First Tamil-First Language Model Trained from Scratch with Tool Calling. A Playbook for Under-Represented Languages | Ramchand Kumaresan, Madhusudhanan G, Kaushik Kezendirane | cs.CL | 2026-10-13 |
| #48 | Do Factuality Methods Make LLMs More Factual? Reproducibility and Fair Comparison | Pratuat Amatya, Vinay Setty, V Venktesh +1 | cs.CL | 2026-10-13 |
| #49 | PRISM: Calibrating Black-Box Judges with Counterfactual Probes | Kakei Yamamoto, So Kuroki | cs.LG | 2026-10-13 |
| #50 | Delegating Earth science model calibration to knowledge-guided LLM agents | Zhanliang Zhu, Ziwei Li, Liujun Zhu +3 | cs.AI | 2026-10-13 |