arXivDaily arXiv每日学术速递 周一至周五更新

AI 大模型

大模型推理能力

大模型数学、逻辑、规划、多步推理和测试时计算能力。

2026-01-13 至 2026-01-13 共收录 7 信号源:cs.CL, cs.AI, cs.LG

1. 逻辑推理 7 篇

2601.07464 2026-01-13 cs.AI 90%

IFDNS: An Iterative Feedback-Driven Neuro-Symbolic Method for Faithful Logical Reasoning

IFDNS: 一种迭代反馈驱动的神经符号方法用于忠实的逻辑推理

Xiaoheng Wang, Tongxuan Liu, Zi Gong, Xianzhe Dong, Yuting Zeng, Minhan Hu, Weizhe Huang, Jing Li

机构 * College of Computer Science and Technology, University of Science and Technology of China(计算机科学与技术学院,中国科学技术大学)

专题命中 逻辑推理 :reasoning(title,abstract);logical reasoning(title,abstract);chain-of-thought(abstract);CoT(abstract)

AI总结 IFDNS通过迭代反馈机制提升LLM逻辑推理的忠实性,有效缓解信息丢失问题,显著提高CoT和CoT-SC的性能。

Comments 13 pages,5 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.07754 2026-01-13 cs.CL 83%

Structure First, Reason Next: Enhancing a Large Language Model using Knowledge Graph for Numerical Reasoning in Financial Documents

先结构,再推理:利用知识图谱增强大型语言模型进行财务文档中的数值推理

Aryan Mishra, Akash Anil

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

AI总结 本文提出利用知识图谱增强大型语言模型,以提高财务文档中的数值推理能力,实验表明该方法在准确性上提升了约12%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.07690 2026-01-13 cs.LO 67%

On Angels and Demons: Strategic (De)Construction of Dynamic Models

关于天使与恶魔:动态模型的战略(破坏)构建

Davide Catta, Rustam Galimullin, Munyque Mittelmann

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

AI总结 本文提出三种逻辑用于研究动态图拓扑中策略的构建与破坏,探讨了其表达能力和模型检查复杂性。

Comments This is an extended version of the paper with the same title that will appear in the proceedings of AAMAS 2026. This version contains a technical appendix with proof details

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.06181 2026-01-13 cs.AI cs.FL cs.LG cs.LO 62%

Neuro-Symbolic Compliance: Integrating LLMs and SMT Solvers for Automated Financial Legal Analysis

神经符号合规:整合大语言模型与SMT求解器用于自动化金融法律分析

Yung-Shen Hsia, Fang Yu, Jie-Hong Roland Jiang

机构 * Department of Management Information Systems, National ChengChi University, Taipei, Taiwan(管理信息系,中华大学,台北,台湾) Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan(电子工程系,台湾大学,台北,台湾)

专题命中 逻辑推理 :reasoning(abstract);分类 cs.AI、cs.LG

AI总结 本研究整合大语言模型与SMT求解器,提出神经符号合规框架,用于自动化金融法律分析,实现形式可验证性和基于优化的合规修正。

Comments 10 pages, 6 tables, 3 figures, accepted by the 2nd ACM AIware Conference

详情

展开后加载摘要…

URL PDF HTML 收藏
2408.17198 2026-01-13 cs.AI cs.LG 62%

Towards Symbolic XAI -- Explanation Through Human Understandable Logical Relationships Between Features

迈向符号化XAI——通过人类可理解的特征间逻辑关系进行解释

Thomas Schnake, Farnoush Rezaei Jafari, Jonas Lederer, Ping Xiong, Shinichi Nakajima, Stefan Gugler, Grégoire Montavon, Klaus-Robert Müller

机构 * Berlin Institute for the Foundations of Learning(柏林学习与数据基础研究所) Machine Learning Group, Technical University of Berlin(柏林技术大学机器学习组) Department of Artificial Intelligence, Korea University(韩国大学人工智能系) Department of Mathematics and Computer Science, Free University of Berlin(柏林自由大学数学与计算机科学系)

专题命中 逻辑推理 :reasoning(abstract);分类 cs.AI、cs.LG

AI总结 本文提出符号化XAI框架,通过人类可理解的逻辑关系解释模型决策,提升AI透明度和可解释性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.06097 2026-01-13 cs.CV cs.AI 57%

Semantic Event Graphs for Long-Form Video Question Answering

语义事件图用于长视频问答

Aradhya Dixit, Tianxi Liang

专题命中 逻辑推理 :reasoning(abstract);分类 cs.AI

AI总结 语义事件图通过轻量级符号接口提升长视频问答效率,减少令牌使用并保持推理能力。

Comments 7 pages, 6 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.05774 2026-01-13 cs.AI 57%

ConstraintLLM: A Neuro-Symbolic Framework for Industrial-Level Constraint Programming

ConstraintLLM: 一种用于工业级约束编程的神经符号框架

Weichun Shi, Minghao Liu, Wanting Zhang, Langchen Shi, Fuqi Jia, Feifei Ma, Jian Zhang

机构 * Hangzhou Institute for Advanced Study, UCAS, Hangzhou, China(杭州高等研究院,UCAS,杭州,中国) University of Oxford, Oxford, UK(牛津大学,牛津,英国) University of Science and Technology Beijing, Beijing, China(北京科技大学,北京,中国) SKLCS and Key Laboratory of System Software, ISCAS, Beijing, China(SKLCS和系统软件重点实验室,ISCAS,北京,中国) Laboratory of Parallel Software and Computational Science, ISCAS, Beijing, China(并行软件与计算科学实验室,ISCAS,北京,中国) University of Chinese Academy of Sciences, Beijing, China(中国科学院大学,北京,中国)

专题命中 逻辑推理 :self-correction(abstract);分类 cs.AI

AI总结 ConstraintLLM是一种专为约束编程设计的神经符号框架,通过引入Constraint-Aware Retrieval Module和Tree-of-Thoughts框架,实现了在工业级约束编程基准上的高性能求解。

Comments Accepted to the 2025 Conference on Empirical Methods in Natural Language Processing (EMNLP 2025), Main Conference

Journal ref Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 15999-16019

详情

展开后加载摘要…

URL PDF HTML 收藏