大模型推理能力

Neural, Symbolic and Neural-Symbolic Reasoning on Knowledge Graphs

Jing Zhang, Bo Chen, Lingxi Zhang, Xirui Ke, Haipeng Ding

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

Comments 29 pages, AI Open Journal 2021

A Logic for Conditional Local Strategic Reasoning

Valentin Goranko, Fengkui Ju

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

Comments 21 pages, to appear in the LORI'2019 special issue of the Journal of Logic, Language and Information

Weakly Supervised Neuro-Symbolic Module Networks for Numerical Reasoning

Amrita Saha, Shafiq Joty, Steven C. H. Hoi

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

Relational-Grid-World: A Novel Relational Reasoning Environment and An Agent Model for Relational Information Extraction

Faruk Kucuksubasi, Elif Surer

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

Neural-Symbolic Computing: An Effective Methodology for Principled Integration of Machine Learning and Reasoning

Artur d'Avila Garcez, Marco Gori, Luis C. Lamb, Luciano Serafini, Michael Spranger, Son N. Tran

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

Automated Reasoning in Deontic Logic

Ulrich Furbach, Claudia Schon, Frieder Stolzenburg

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

Journal ref In M. Narasimha Murty, Xiangjian He, Raghavendra Rao Chillarige, and Paul Weng, editors, Proc. of MIWAI 2014: Multi-disciplinary International Workshop on Artificial Intelligence, LNAI 8875, pp. 57-68, Bangalore, India, 2014. Springer

The Complexity of Limited Belief Reasoning -- The Quantifier-Free Case

Yijia Chen, Abdallah Saffidine, Christoph Schwering

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

Comments 15 pages, 1 figure, 1 table, Twenty-seventh International Joint Conference on Artificial Intelligence (IJCAI-18)

Reasoning about Unforeseen Possibilities During Policy Learning

Craig Innes, Alex Lascarides, Stefano V Albrecht, Subramanian Ramamoorthy, Benjamin Rosman

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

Practical Reasoning with Norms for Autonomous Software Agents (Full Edition)

Zohreh Shams, Marina De Vos, Julian Padget, Wamberto W. Vasconcelos

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

Towards common-sense reasoning via conditional simulation: legacies of Turing in Artificial Intelligence

Cameron E. Freer, Daniel M. Roy, Joshua B. Tenenbaum

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

Comments 51 pages, 6 figures, 1 table. Slight typographical fixes

Journal ref Turing's Legacy: Developments from Turing's Ideas in Logic, ed. Rod Downey, ASL Lecture Notes in Logic 42, Cambridge University Press, 2014

Guarantees and Limits of Preprocessing in Constraint Satisfaction and Reasoning

Serge Gaspers, Stefan Szeider

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

Comments arXiv admin note: substantial text overlap with arXiv:1104.2541, arXiv:1104.5566

Reasoning About the Transfer of Control

Wiebe van der Hoek, Dirk Walther, Michael Wooldridge

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

Journal ref Journal Of Artificial Intelligence Research, Volume 37, pages 437-477, 2010

An Integrated Framework for Learning and Reasoning

C. G. Giraud-Carrier, T. R. Martinez

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

Comments See http://www.jair.org/ for an online appendix and other files accompanying this article

Journal ref Journal of Artificial Intelligence Research, Vol 3, (1995), 147-185

Combining Proofs for Description Logic and Concrete Domain Reasoning (Technical Report)

Christian Alrabbaa, Franz Baader, Stefan Borgwardt, Patrick Koopmann, Alisa Kovtunova

专题命中 逻辑推理 ：reasoning(title,abstract)

Comments Extended version of a paper accepted at the 7th International Joint Conference on Rules and Reasoning (RuleML+RR 2023). The software and the dataset described in the paper are available online, see https://doi.org/10.5281/zenodo.8208780

How well do SOTA legal reasoning models support abductive reasoning?

Ha-Thanh Nguyen, Randy Goebel, Francesca Toni, Kostas Stathis, Ken Satoh

专题命中 逻辑推理 ：reasoning(title,abstract)

Comments Workshop on Logic Programming and Legal Reasoning, @ICLP 2023

WCM: World-Cognition Model for Generalizable Human-Robot Interaction

WCM：用于通用人机交互的世界认知模型

Yuzhen Chen, KC Zhou

专题命中 逻辑推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.AI、cs.LG

AI总结 针对机器人在物理任务交互困难的问题，提出基于SLAK架构和异步运行时的世界认知模型（WCM），引入人在回路教学模式，经思维链监督改进模型，在九个现实世界人机交互任务中平均成功率达73.8%。

More Yap Less Meaning: Uncovering Self-Improvement Behavior in SLMs

更多废话，更少意义：揭示小语言模型中的自我改进行为

Marina Igitkhanian, Erik Arakelyan

机构 * American University of Armenia（亚美尼亚美国大学） ； NVIDIA（英伟达） 纠错

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

AI总结 本研究通过构建充分性测试，发现小语言模型在自我纠正中仅获得4.4%的准确率提升，且较长的提示反而与错误答案正相关，表明其推理能力有限。

Comments GEM Workshop at ACL 2026

Robustness of Prompting: Enhancing Robustness of Large Language Models Against Prompting Attacks

提示的鲁棒性：增强大型语言模型对抗提示攻击的鲁棒性

Lin Mu, Guowei Chu, Li Ni, Lei Sang, Yiwen Zhang

机构 * School of Computer Science and Technology, Anhui University（安徽大学计算机科学与技术学院） 纠错

专题命中 逻辑推理 ：reasoning(abstract)；chain-of-thought(abstract)；logical reasoning(abstract)；分类 cs.CL、cs.AI

AI总结 提出RoP（提示鲁棒性）策略，通过错误校正和引导两个阶段，增强LLM对输入扰动的鲁棒性，在算术、常识和逻辑推理任务上显著提升性能。

Comments Accepted by IEEE Transactions on Artificial Intelligence

MERIT: Memory-Enhanced Retrieval for Interpretable Knowledge Tracing

MERIT: 用于可解释知识追踪的记忆增强检索

Runze Li, Kedi Chen, Guwei Feng, Mo Yu, Jun Wang, Wei Zhang

机构 * School of Computer Science and Technology, East China Normal University（东华师范大学计算机科学与技术学院） ； Shanghai Innovation Institute（上海创新研究院） ； WeChat AI, Tencent（微信AI，腾讯） 纠错

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

AI总结 MERIT提出一种无需训练的框架，结合冻结LLM推理与结构化教学记忆，通过语义去噪将学生分类为潜在认知模式，并构建模式库生成显式推理依据，提升教育诊断的可解释性和效率。

A Balanced Neuro-Symbolic Approach for Commonsense Abductive Logic

一种平衡的神经符号方法用于常识归纳逻辑

Joseph Cotnareanu, Didier Chetelat, Yingxue Zhang, Mark Coates

机构 * McGill University（麦吉尔大学） ； Huawei Noah’s Ark Lab（华为诺亚实验室） ； International Laboratory on Learning Systems (ILLS)（学习系统国际实验室） ； MILA（蒙特利尔人工智能实验室） 纠错

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

AI总结 本文提出一种平衡神经和符号方法，通过迭代增强逻辑问题中的常识关系，提升复杂推理任务的性能。

Digital Metabolism: Decoupling Logic from Facts via Regenerative Unlearning -- Towards a Pure Neural Logic Core

数字代谢：通过再生反学习解耦逻辑与事实 -- 向纯神经逻辑核心迈进

Mengmeng Peng, Zhenyu Fang, He Sun

专题命中 逻辑推理 ：reasoning(abstract)；chain-of-thought(abstract)；CoT(abstract)；分类 cs.AI、cs.LG

AI总结 本文提出'数字代谢'假说，通过再生反学习解耦逻辑与事实，实现纯神经逻辑核心，提升模型推理能力并减少幻觉。

GroupDebate: Enhancing the Efficiency of Multi-Agent Debate Using Group Discussion

GroupDebate: 通过群体讨论提升多智能体辩论效率

Tongxuan Liu, Xingyu Wang, Weizhe Huang, Wenjiang Xu, Yuting Zeng, Lei Jiang, Hailong Yang, Jing Li

机构 * University of Science and Technology of China（中国科学技术大学） ； Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所） ； Beihang University（北航） 纠错

专题命中 逻辑推理 ：reasoning(abstract)；chain-of-thought(abstract)；logical reasoning(abstract)；分类 cs.CL、cs.AI

AI总结 本文提出GroupDebate方法，通过将智能体划分为小组并共享中间结果，减少多智能体辩论的token成本，提升效率和准确性。

Comments Accepted by AAMAS 2026

DiLA: Enhancing LLM Tool Learning with Differential Logic Layer

Yu Zhang, Hui-Ling Zhen, Zehua Pei, Yingzhao Lian, Lihao Yin, Mingxuan Yuan, Bei Yu

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

Comments arXiv admin note: text overlap with arXiv:2305.12295 by other authors

Integrating Expert Knowledge into Logical Programs via LLMs

Franciszek Górski, Oskar Wysocki, Marco Valentino, Andre Freitas

机构 * Multimedia Systems Department, Gdansk University of Technology, Poland（格但斯克技术大学多媒体系统系，波兰） ； Department of Computer Science, University of Manchester, United Kingdom（曼彻斯特大学计算机科学系，英国） ； National Biomarker Centre (NBC), CRUK Manchester Institute, United Kingdom（英国CRUK曼彻斯特研究所国家生物标记中心） ； Idiap Research Institute, Martigny, Switzerland（瑞士马尔蒂尼Idiap研究所） 纠错

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

InfiGUI-R1: Advancing Multimodal GUI Agents from Reactive Actors to Deliberative Reasoners

Yuhang Liu, Pengxiang Li, Congkai Xie, Xavier Hu, Xiaotian Han, Shengyu Zhang, Hongxia Yang, Fei Wu

机构 * Zhejiang University（浙江大学） ； Dalian University of Technology（大连理工大学） ； Reallm Labs（Reallm实验室） ； The Hong Kong Polytechnic University（香港理工大学） 纠错

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

Comments 10 pages, 3 figures, work in progress

On the logical skills of large language models: evaluations using arbitrarily complex first-order logic problems

Shokhrukh Ibragimov, Arnulf Jentzen, Benno Kuckuck

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

Comments 67 pages, 24 figures

Logic Agent: Enhancing Validity with Logic Rule Invocation

Hanmeng Liu, Zhiyang Teng, Chaoli Zhang, Yue Zhang

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

Comments The experiment is subject to certain errors

Artificial General Intelligence (AGI)-Native Wireless Systems: A Journey Beyond 6G

Walid Saad, Omar Hashash, Christo Kurisummoottil Thomas, Christina Chaccour, Merouane Debbah, Narayan Mandayam, Zhu Han

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

Neurosymbolic AI for Reasoning on Biomedical Knowledge Graphs

Lauren Nicole DeLong, Ramon Fernández Mir, Zonglin Ji, Fiona Niamh Coulter Smith, Jacques D. Fleuriot

专题命中 逻辑推理 ：reasoning(title,comments)；分类 cs.AI、cs.LG；logical reasoning(comments)

Comments Proceedings of the $\mathit{40}^{th}$ International Conference on Machine Learning: Workshop on Knowledge and Logical Reasoning in the Era of Data-driven Learning (https://klr-icml2023.github.io/schedule.html). PMLR 202, 2023. Condensed, workshop-ready version of previous survey, arXiv:2302.07200 , which is under review. 13 pages (9 content, 4 references), 3 figures, 1 table

Harnessing the Power of Large Language Models for Natural Language to First-Order Logic Translation

Yuan Yang, Siheng Xiong, Ali Payani, Ehsan Shareghi, Faramarz Fekri

专题命中 逻辑推理 ：reasoning(abstract)；chain-of-thought(abstract)；verifier(abstract)；分类 cs.CL、cs.AI