大模型推理能力

Reasoning Model Unlearning: Forgetting Traces, Not Just Answers, While Preserving Reasoning Skills

Changsheng Wang, Chongyu Fan, Yihua Zhang, Jinghan Jia, Dennis Wei, Parikshit Ram, Nathalie Baracaldo, Sijia Liu

机构 * Michigan State University（密歇根州立大学） ； IBM Research（IBM研究院） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.AI、cs.LG

Comments Accepted by EMNLP 2025

Thinking with Nothinking Calibration: A New In-Context Learning Paradigm in Reasoning Large Language Models

Haotian Wu, Bo Xu, Yao Shu, Menglin Yang, Chengwei Qin

机构 * The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州）） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL

Path Drift in Large Reasoning Models:How First-Person Commitments Override Safety

Yuyi Huang, Runzhe Zhan, Lidia S. Chao, Ailin Tao, Derek F. Wong

机构 * The Second Affiliated Hospital, Guangdong Provincial Key Laboratory of Allergy and Clinical Immunology, Guangzhou Medical University（广东省过敏与临床免疫学重点实验室，广州医科大学第二附属医院） ； NLP 2 CT Lab, Department of Computer and Information Science, University of Macau（澳门大学计算机与信息科学系NLP2CT实验室） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL

Dyna-Think: Synergizing Reasoning, Acting, and World Model Simulation in AI Agents

Xiao Yu, Baolin Peng, Ruize Xu, Michel Galley, Hao Cheng, Suman Nath, Jianfeng Gao, Zhou Yu

机构 * Columbia University（哥伦比亚大学） ； Microsoft Research（微软研究院） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；planning(abstract)；分类 cs.CL、cs.AI、cs.LG

Hound: Relation-First Knowledge Graphs for Complex-System Reasoning in Security Audits

Bernhard Mueller

机构 * Bernhard Mueller（伯纳德·穆勒） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；planning(abstract)；分类 cs.AI、cs.LG

ARM: Adaptive Reasoning Model

Siye Wu, Jian Xie, Yikai Zhang, Aili Chen, Kai Zhang, Yu Su, Yanghua Xiao

机构 * Shanghai Key Laboratory of Data Science, College of Computer Science and Artificial Intelligence, Fudan University（上海数据科学关键实验室，计算机科学与人工智能学院，复旦大学） ； The Ohio State University（俄亥俄州立大学） ； Shanghai Academy of AI for Science（上海人工智能科学研究院） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；CoT(abstract)；分类 cs.CL

Comments NeurIPS 2025 (Spotlight)

J1: Incentivizing Thinking in LLM-as-a-Judge via Reinforcement Learning

Chenxi Whitehouse, Tianlu Wang, Ping Yu, Xian Li, Jason Weston, Ilia Kulikov, Swarnadeep Saha

机构 * FAIR at Meta（Meta 公司） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；chain-of-thought(abstract)；self-correction(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 10 pages, 13 tables, 14 figures

Information-Preserving Reformulation of Reasoning Traces for Antidistillation

Jiayu Ding, Lei Cui, Li Dong, Nanning Zheng, Furu Wei

机构 * IAIR, Xi’an Jiaotong University（人工智能研究院，西安交通大学） ； Microsoft Research（微软研究院） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；分类 cs.CL

From <Answer> to <Think>: Multidimensional Supervision of Reasoning Process for LLM Optimization

Beining Wang, Weihang Su, Hongtao Tian, Tao Yang, Yujia Zhou, Ting Yao, Qingyao Ai, Yiqun Liu

机构 * Tsinghua University（清华大学） ； Tencent Inc（腾讯公司） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)；分类 cs.AI

Real-Time Health Analytics Using Ontology-Driven Complex Event Processing and LLM Reasoning: A Tuberculosis Case Study

Ritesh Chandra, Sonali Agarwal, Navjot Singh

专题命中 复杂问题求解 ：reasoning(title,abstract)；分类 cs.AI

Comments 14 table. 20 figure

The Idola Tribus of AI: Large Language Models tend to perceive order where none exists

Shin-nosuke Ishikawa, Masato Todo, Taiki Ogihara, Hirotsugu Ohba

机构 * Graduate School of Artificial Intelligence and Science, Rikkyo University（人工智能与科学研究生院，立命馆大学） ； Strategic Digital Business Unit, Mamezou Co., Ltd.（战略数字商务部，麦迈株式会社） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；chain-of-thought(abstract)；logical reasoning(abstract)；分类 cs.CL、cs.AI

Comments 14 pages, 3 figures, accepted to Findings of EMNLP 2025

COCO-Tree: Compositional Hierarchical Concept Trees for Enhanced Reasoning in Vision Language Models

Sanchit Sinha, Guangzhi Xiong, Aidong Zhang

机构 * University of Virginia（弗吉尼亚大学） 纠错

专题命中 复杂问题求解 ：reasoning(title,abstract)

Comments EMNLP 2025 (main)

Analyzing and Internalizing Complex Policy Documents for LLM Agents

Jiateng Liu, Zhenhailong Wang, Xiaojiang Huang, Yingjie Li, Xing Fan, Xiang Li, Chenlei Guo, Ruhi Sarikaya, Heng Ji

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Amazon（亚马逊） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.AI

Comments 42 pages

Do LLMs Really Need 10+ Thoughts for "Find the Time 1000 Days Later"? Towards Structural Understanding of LLM Overthinking

Xinliang Frederick Zhang, Anhad Mohananey, Alexandra Chronopoulou, Pinelopi Papalampidi, Somit Gupta, Tsendsuren Munkhdalai, Lu Wang, Shyam Upadhyay

机构 * University of Michigan（密歇根大学） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL

Comments 30 pages, 41 figures, 10 tables. Preprint

ProSEA: Problem Solving via Exploration Agents

William Nguyen, Vinh Luong, Christopher Nguyen

机构 * Aitomatic, Inc.（Aitomatic公司） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；planning(abstract)；分类 cs.AI

PolyVer: A Compositional Approach for Polyglot System Modeling and Verification

Pei-Wei Chen, Shaokai Lin, Adwait Godbole, Ramneet Singh, Elizabeth Polgreen, Edward A. Lee, Sanjit A. Seshia

专题命中 复杂问题求解 ：reasoning(abstract)；verifier(abstract)

Comments Proceedings of the 25th Conference on Formal Methods in Computer-Aided Design (FMCAD 2025)

Discursive Circuits: How Do Language Models Understand Discourse Relations?

Yisong Miao, Min-Yen Kan

机构 * Web IR / NLP Group (WING)（网络信息检索与自然语言处理组） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.CL、cs.LG

Comments Accepted to EMNLP 2025 (Main Conference); 9 pages, 8 figures, 5 tables (20 pages, 12 figures, 14 tables including references and appendices)

CALM: A Causal Analysis Language Model for Tabular Data in Complex Systems with Local Scores, Conditional Independence Tests, and Relation Attributes

Zhenjiang Fan, Zengyi Qin, Yuanning Zheng, Bo Xiong, Summer Han

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.AI、cs.LG

Preserving LLM Capabilities through Calibration Data Curation: From Analysis to Optimization

Bowei He, Lihao Yin, Huiling Zhen, Shuqi Liu, Han Wu, Xiaokun Zhang, Mingxuan Yuan, Chen Ma

机构 * Department of Computer Science, City University of Hong Kong（香港城市大学计算机科学系） ； Huawei（华为） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.CL

Comments Accepted by NeurIPS 2025

LIDDIA: Language-based Intelligent Drug Discovery Agent

Reza Averly, Frazier N. Baker, Ian A. Watson, Xia Ning

机构 * Department of Computer Science and Engineering, The Ohio State University, USA（俄亥俄州立大学计算机科学与工程系） ； Independent Researcher（独立研究者） ； Department of Biomedical Informatics, The Ohio State University, USA（俄亥俄州立大学生物医学信息学系） ； Translational Data Analytics Institute, The Ohio State University, USA（俄亥俄州立大学转化数据分析研究所） ； College of Pharmacy, The Ohio State University, USA（俄亥俄州立大学药学院） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.CL

Comments EMNLP 2025 Main Conference

Bridging Semantics & Structure for Software Vulnerability Detection using Hybrid Network Models

Jugal Gajjar, Kaustik Ranaware, Kamalasankari Subramaniakuppusamy

机构 * The George Washington University（乔治·华盛顿大学） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.AI

Comments 13 pages, 3 figures, 5 tables, 14 equations, accepted at the 14th International Conference on Complex Networks and Their Applications (COMPLEX NETWORKS 2025) and the conference proceedings will be published by Springer in the Studies in Computational Intelligence series

Don't Throw Away Your Pretrained Model

Shangbin Feng, Wenhao Yu, Yike Wang, Hongming Zhang, Yulia Tsvetkov, Dong Yu

机构 * University of Washington（华盛顿大学） ； Tencent AI Seattle Lab（腾讯AI西雅图实验室） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)；分类 cs.CL

LLM-HBT: Dynamic Behavior Tree Construction for Adaptive Coordination in Heterogeneous Robots

Chaoran Wang, Jingyuan Sun, Yanhui Zhang, Mingyu Zhang, Changju Wu

机构 * School of Aeronautic and Astronautics, Zhejiang University（浙江大学航空宇航学院） ； Shanghai Huawei Technologies Co., Ltd.（上海华为技术有限公司） 纠错

专题命中 复杂问题求解 ：reasoning(abstract)

Comments It contains 8 pages, 7 figures and 4 tables. This paper is submitted to ICRA 2026