大模型推理能力

Layer Importance for Mathematical Reasoning is Forged in Pre-Training and Invariant after Post-Training

Aadim Nepal, Safal Shrestha, Anubhav Shrestha, Minwu Kim, Jalal Naghiyev, Ravid Shwartz-Ziv, Keith Ross

机构 * New York University Abu Dhabi（纽约大学阿布扎赫尔分校） ； Technical University of Munich（慕尼黑技术大学） ； NYU Center for Data Science（纽约大学数据科学中心） 纠错

专题命中 数学推理 ：reasoning(title,abstract)；math reasoning(abstract)；分类 cs.AI、cs.LG

R2R: Efficiently Navigating Divergent Reasoning Paths with Small-Large Model Token Routing

Tianyu Fu, Yi Ge, Yichen You, Enshu Liu, Zhihang Yuan, Guohao Dai, Shengen Yan, Huazhong Yang, Yu Wang

机构 * Tsinghua University（清华大学） ； Infinigence AI ； Shanghai Jiao Tong University（上海交通大学） 纠错

专题命中 数学推理 ：reasoning(title,abstract)；分类 cs.CL、cs.AI、cs.LG

Distilling LLM Agent into Small Models with Retrieval and Code Tools

Minki Kang, Jongwon Jeong, Seanie Lee, Jaewoong Cho, Sung Ju Hwang

机构 * KAIST（韩国科学技术院） ； University of Wisconsin-Madison（威斯康星大学麦迪逊分校） ； KRAFTON（KRAFTON公司） 纠错

专题命中 数学推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.CL、cs.AI

Comments NeurIPS 2025 Spotlight

LiveTradeBench: Seeking Real-World Alpha with Large Language Models

Haofei Yu, Fenghai Li, Jiaxuan You

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

专题命中 数学推理 ：reasoning(abstract)；math reasoning(abstract)；分类 cs.CL、cs.AI

Comments 16 pages

PDE-SHARP: PDE Solver Hybrids through Analysis and Refinement Passes

Shaghayegh Fazliani, Madeleine Udell

机构 * Department of Mathematics Stanford University（数学系 斯坦福大学） ； Department of Management Science & Engineering Stanford University（管理科学与工程系 斯坦福大学） 纠错

专题命中 数学推理 ：reasoning(abstract)；chain-of-thought(abstract)；分类 cs.LG

Why Less is More (Sometimes): A Theory of Data Curation

Elvis Dohmatob, Mohammad Pezeshki, Reyhane Askari-Hemmat

机构 * Concordia University（康科德大学） ； FAIR at Meta（Meta的FAIR） ； Mila--Quebec AI Institute（魁北克AI研究所） 纠错

专题命中 数学推理 ：reasoning(abstract)；分类 cs.LG

Data Dependency-Aware Code Generation from Enhanced UML Sequence Diagrams

Wenxin Mao, Zhitao Wang, Long Wang, Sirong Chen, Cuiyun Gao, Luyang Cao, Ziming Liu, Qiming Zhang, Jun Zhou, Zhi Jin

机构 * WeChat Pay, Tencent Inc（腾讯公司） ； The Chinese University of Hong Kong（香港中文大学） ； Wuhan University（武汉大学） 纠错

专题命中 数学推理 ：reasoning(abstract)；分类 cs.AI

miniF2F-Lean Revisited: Reviewing Limitations and Charting a Path Forward

Azim Ospanov, Farzan Farnia, Roozbeh Yousefzadeh

机构 * Huawei Hong Kong Research Center（华为香港研究中心） ； Department of Computer Science & Engineering, The Chinese University of Hong Kong（香港中文大学计算机科学与工程系） 纠错

专题命中 代码与定理证明 ：reasoning(abstract)；分类 cs.AI

LFC-DA: Logical Formula-Controlled Data Augmentation for Enhanced Logical Reasoning

Shenghao Li

机构 * Guangzhou University（广州大学） 纠错

专题命中 逻辑推理 ：reasoning(title,abstract)；logical reasoning(title)；分类 cs.CL

Comments 10 pages, 6 figures

U2F: Encouraging SWE-Agent to Seize Novelty without Losing Feasibility

Wencheng Ye, Yan Liu

专题命中 逻辑推理 ：reasoning(abstract)；logical reasoning(abstract)

PlanU: Large Language Model Reasoning through Planning under Uncertainty

Ziwei Deng, Mian Deng, Chenjing Liang, Zeming Gao, Chennan Ma, Chenxing Lin, Haipeng Zhang, Songzhu Mei, Siqi Shen, Cheng Wang

机构 * Fujian Key Laboratory of Sensing and Computing for Smart Cities, School of Informatics, Xiamen University（福建智慧城市感知与计算重点实验室，信息学院，厦门大学） ； Key Laboratory of Multimedia Trusted Perception and Efficient Computing, XMU（多媒体可信感知与高效计算重点实验室，XMU） ； School of Computer, National University of Defense Technology（计算机学院，国防科技大学） 纠错

专题命中 规划推理 ：reasoning(title,abstract)；planning(title,abstract)；分类 cs.AI

Comments 38 pages, 19 figures, NeurIPS 2025 Accepted

Let the Bees Find the Weak Spots: A Path Planning Perspective on Multi-Turn Jailbreak Attacks against LLMs

Yize Liu, Yunyun Hou, Aina Sui

专题命中 规划推理 ：planning(title,abstract)；分类 cs.CL

Incorporating Quality of Life in Climate Adaptation Planning via Reinforcement Learning

Miguel Costa, Arthur Vandervoort, Martin Drews, Karyn Morrissey, Francisco C. Pereira

机构 * Technical University of Denmark（丹麦技术大学） ； Univ. of Galway（Galway大学） 纠错

专题命中 规划推理 ：planning(title,abstract)；分类 cs.LG

Comments Accepted for presentation at AI in Science (AIS) 2025

Multi-Objective Planning with Contextual Lexicographic Reward Preferences

Pulkit Rustagi, Yashwanthi Anand, Sandhya Saisubramanian

机构 * Oregon State University（俄勒冈州立大学） 纠错

专题命中 规划推理 ：planning(title,abstract)；分类 cs.AI

Comments 9 pages, 5 figures, 2 tables, To appear in Proceedings of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS) 2025

ALAS: Transactional and Dynamic Multi-Agent LLM Planning

Longling Geng, Edward Y. Chang

专题命中 规划推理 ：planning(title,abstract)

Digital Transformation Chatbot (DTchatbot): Integrating Large Language Model-based Chatbot in Acquiring Digital Transformation Needs

Jiawei Zheng, Gokcen Yilmaz, Ji Han, Saeema Ahmed-Kristensen

机构 * DigitLab, University of Exeter, UK（DigitLab，埃克塞特大学，英国） 纠错

专题命中 规划推理 ：reasoning(abstract)；planning(abstract)；分类 cs.AI

Comments Accepted by the International Conference on Human-Computer Interaction

Reinforcement Learning Foundations for Deep Research Systems: A Survey

Wenjun Li, Zhi Chen, Jingru Lin, Hannan Cao, Wei Han, Sheng Liang, Zhi Zhang, Kuicai Dong, Dexun Li, Chen Zhang, Yong Liu

专题命中 规划推理 ：reasoning(abstract)；分类 cs.CL、cs.AI

Comments 39 pages, second version

Kosmos: An AI Scientist for Autonomous Discovery

Ludovico Mitchener, Angela Yiu, Benjamin Chang, Mathieu Bourdenx, Tyler Nadolski, Arvis Sulovari, Eric C. Landsness, Daniel L. Barabasi, Siddharth Narayanan, Nicky Evans, Shriya Reddy, Martha Foiani, Aizad Kamal, Leah P. Shriver, Fang Cao, Asmamaw T. Wassie, Jon M. Laurent, Edwin Melville-Green, Mayk Caldas, Albert Bou, Kaleigh F. Roberts, Sladjana Zagorac, Timothy C. Orr, Miranda E. Orr, Kevin J. Zwezdaryk, Ali E. Ghareeb, Laurie McCoy, Bruna Gomes, Euan A. Ashley, Karen E. Duff, Tonio Buonassisi, Tom Rainforth, Randall J. Bateman, Michael Skarlinski, Samuel G. Rodriques, Michaela M. Hinks, Andrew D. White

专题命中 规划推理 ：reasoning(abstract)；分类 cs.AI

Comments Revision: figure layout changes and minor text edits

Divide by Question, Conquer by Agent: SPLIT-RAG with Question-Driven Graph Partitioning

Ruiyi Yang, Hao Xue, Imran Razzak, Shirui Pan, Hakim Hacid, Flora D. Salim

机构 * University of New South Wales（新南威尔士大学） ； Mohamed Bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学） ； Griffith University（格里菲斯大学） ； Technology Innovation Institute（技术创新研究所） 纠错

专题命中 规划推理 ：reasoning(abstract)；分类 cs.AI

Comments 20 pages, 4 figures

Digital Twin-Driven Pavement Health Monitoring and Maintenance Optimization Using Graph Neural Networks

Mohsin Mahmud Topu, Mahfuz Ahmed Anik, Azmine Toushik Wasi, Md Manjurul Ahsan

专题命中 规划推理 ：planning(abstract)；分类 cs.LG

On Improvisation and Open-Endedness: Insights for Experiential AI

Botao 'Amber' Hu

专题命中 规划推理 ：planning(abstract)；分类 cs.AI

Comments Submitted to AAAI 2026 Creative AI for Live Interactive Performances Workshop (CLIP) as a work-in-progress paper

Agentic Meta-Orchestrator for Multi-task Copilots

Xiaofeng Zhu, Yunshen Zhou

机构 * Microsoft Corporation（微软公司） 纠错

专题命中 规划推理 ：planning(abstract)；分类 cs.AI

Journal ref ICDM RARA workshop 2025

Multi-Agent Reinforcement Learning for Autonomous Multi-Satellite Earth Observation: A Realistic Case Study

Mohamad A. Hady, Siyi Hu, Mahardhika Pratama, Jimmy Cao, Ryszard Kowalczyk

机构 * University of South Australia（南澳大利亚大学） ； Systems Research Institute, Polish Academy of Sciences（波兰科学院系统研究所） 纠错

专题命中 规划推理 ：planning(abstract)；分类 cs.AI

SENT Map -- Semantically Enhanced Topological Maps with Foundation Models

Raj Surya Rajendran Kathirvel, Zach A Chavis, Stephen J. Guy, Karthik Desingh

机构 * Minnesota Robotics Institute (MnRI)（明尼苏达州机器人研究所） ； Department of Computer Science and Engineering (CS&E)（计算机科学与工程系） ； University of Minnesota（明尼苏达大学） 纠错

专题命中 规划推理 ：planning(abstract)

Comments Accepted at ICRA 2025 Workshop on Foundation Models and Neuro-Symbolic AI for Robotics

Domain-Adaptive Transformer for Data-Efficient Glioma Segmentation in Sub-Saharan MRI

Ilerioluwakiiye Abolade, Aniekan Udo, Augustine Ojo, Abdulbasit Oyetunji, Hammed Ajigbotosho, Aondana Iorumbur, Confidence Raymond, Maruf Adewole

机构 * Federal University of Agriculture Abeokuta（非洲联邦农业大学阿贝奥克塔） 纠错

专题命中 规划推理 ：planning(abstract)

Comments 4 pages, 2 figures. Accepted as an abstract at the Women in Machine Learning (WiML) Workshop at NeurIPS 2025

Struct2D: A Perception-Guided Framework for Spatial Reasoning in MLLMs

Fangrui Zhu, Hanhui Wang, Yiming Xie, Jing Gu, Tianye Ding, Jianwei Yang, Huaizu Jiang

机构 * Northeastern University（东北大学） ； Microsoft Research（微软研究院） ； University of Southern California（南加州大学） ； University of California, Santa Cruz（加州大学圣克鲁兹分校） 纠错

专题命中 视觉空间推理 ：reasoning(title,abstract)；planning(abstract)

Comments NeurIPS 2025, code link: https://github.com/neu-vi/struct2d

DiffSpectra: Molecular Structure Elucidation from Spectra using Diffusion Models

Liang Wang, Yu Rong, Tingyang Xu, Zhenyi Zhong, Zhiyuan Liu, Pengju Wang, Deli Zhao, Qiang Liu, Shu Wu, Liang Wang, Yang Zhang

机构 * NLPR, MAIS, Institute of Automation（NLPR、MAIS、自动化研究所） ； School of Artificial Intelligence（人工智能学院） ； University of Chinese Academy of Sciences（中国科学院大学） ； Alibaba Group（阿里巴巴集团） ； Hupan Lab（华普实验室） ； College of Intelligence and Computing（智能与计算学院） ； National University of Singapore（新加坡国立大学） ； Cancer Science Institute of Singapore（新加坡癌症科学研究所） 纠错

专题命中 视觉空间推理 ：reasoning(abstract)；分类 cs.AI、cs.LG

Read Your Own Mind: Reasoning Helps Surface Self-Confidence Signals in LLMs

Jakub Podolak, Rajeev Verma

机构 * University of Amsterdam（阿姆斯特丹大学） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；chain-of-thought(abstract)；test-time compute(abstract)；分类 cs.CL

Comments Presented at UncertaiNLP Workshop at EMNLP 2025 https://aclanthology.org/2025.uncertainlp-main.21.pdf

Journal ref UncertaiNLP Workshop at Empirical Methods in Natural Language Processing 2025 (EMNLP 2025)

PhysicsEval: Inference-Time Techniques to Improve the Reasoning Proficiency of Large Language Models on Physics Problems

Oshayer Siddique, J. M Areeb Uzair Alam, Md Jobayer Rahman Rafy, Syed Rifat Raiyan, Hasan Mahmud, Md Kamrul Hasan

机构 * Systems and Software Lab (SSL) Department of Computer Science and Engineering（系统与软件实验室（SSL）计算机科学与工程系） 纠错

专题命中 测试时计算 ：reasoning(title,abstract)；分类 cs.CL、cs.AI

Comments Accepted in Findings of the Association for Computational Linguistics: IJCNLP-AACL 2025, 23 pages, 4 figures, 8 tables

Voost: A Unified and Scalable Diffusion Transformer for Bidirectional Virtual Try-On and Try-Off

Seungyong Lee, Jeong-gi Kwak

机构 * NXN Labs（NXN实验室） ； University of British Columbia（不列颠哥伦比亚大学） 纠错

专题命中 测试时计算 ：reasoning(abstract)；分类 cs.AI、cs.LG

Comments Accepted to SIGGRAPH Asia 2025, project page: https://nxnai.github.io/Voost/