大模型推理能力

LEAN-GitHub: Compiling GitHub LEAN repositories for a versatile LEAN prover

Zijian Wu, Jiayu Wang, Dahua Lin, Kai Chen

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

Disentangled Representations for Causal Cognition

Filippo Torresan, Manuel Baltieri

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

Comments 49 pages, 9 figures

Eliciting Problem Specifications via Large Language Models

Robert E. Wray, James R. Kirk, John E. Laird

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

Comments 18 pages, Appendix. Revised in response to reviewer feedback. Accepted for Advances in Cognitive Systems (Jun 2024, Palermo)

Multi-hop Question Answering

Vaibhav Mavi, Anubhav Jangra, Adam Jatowt

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

Comments Published at Foundations and Trends in Information Retrieval

HALO: An Ontology for Representing and Categorizing Hallucinations in Large Language Models

Navapat Nananukul, Mayank Kejriwal

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

Comments This paper has been accepted and orally presented in "SPIE Defense + Commercial Sensing (DCS 2024)" in National Harbor, Maryland, April 2024

Fine-Tuning Language Models Using Formal Methods Feedback

Yunhao Yang, Neel P. Bhatt, Tyler Ingebrand, William Ward, Steven Carr, Zhangyang Wang, Ufuk Topcu

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

Magnushammer: A Transformer-Based Approach to Premise Selection

Maciej Mikuła, Szymon Tworkowski, Szymon Antoniak, Bartosz Piotrowski, Albert Qiaochu Jiang, Jin Peng Zhou, Christian Szegedy, Łukasz Kuciński, Piotr Miłoś, Yuhuai Wu

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

Comments ICLR 2024

SwissNYF: Tool Grounded LLM Agents for Black Box Setting

Somnath Sendhil Kumar, Dhruv Jain, Eshaan Agarwal, Raunak Pandey

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

Anytime Approximate Formal Feature Attribution

Jinqiang Yu, Graham Farr, Alexey Ignatiev, Peter J. Stuckey

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

Towards Concept-Aware Large Language Models

Chen Shani, Jilles Vreeken, Dafna Shahaf

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

Comments EMNLP 2023 findings long paper

Language Models with Rationality

Nora Kassner, Oyvind Tafjord, Ashish Sabharwal, Kyle Richardson, Hinrich Schuetze, Peter Clark

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

Large Language Model for Science: A Study on P vs. NP

Qingxiu Dong, Li Dong, Ke Xu, Guangyan Zhou, Yaru Hao, Zhifang Sui, Furu Wei

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

Comments 73 pages

Hardness of Deceptive Certificate Selection

Stephan Wäldchen

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

Comments 15 pages, 3 figures

Leveraging Large Language Models in Conversational Recommender Systems

Luke Friedman, Sameer Ahuja, David Allen, Zhenning Tan, Hakim Sidahmed, Changbo Long, Jun Xie, Gabriel Schubiner, Ajay Patel, Harsh Lara, Brian Chu, Zexi Chen, Manoj Tiwari

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

Draft, Sketch, and Prove: Guiding Formal Theorem Provers with Informal Proofs

Albert Q. Jiang, Sean Welleck, Jin Peng Zhou, Wenda Li, Jiacheng Liu, Mateja Jamnik, Timothée Lacroix, Yuhuai Wu, Guillaume Lample

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

Exploring Length Generalization in Large Language Models

Cem Anil, Yuhuai Wu, Anders Andreassen, Aitor Lewkowycz, Vedant Misra, Vinay Ramasesh, Ambrose Slone, Guy Gur-Ari, Ethan Dyer, Behnam Neyshabur

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

NaturalProver: Grounded Mathematical Proof Generation with Language Models

Sean Welleck, Jiacheng Liu, Ximing Lu, Hannaneh Hajishirzi, Yejin Choi

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

Comments NeurIPS 2022

Home Run: Finding Your Way Home by Imagining Trajectories

Daria de Tinguy, Pietro Mazzaglia, Tim Verbelen, Bart Dhoedt

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

Comments accepted for International Workshop on Active Inference IWAI 2022, ECML PKDD workshop

Natural language understanding for logical games

Adrian Groza, Cristian Nitu

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

Formal Methods for the Informal Engineer: Workshop Recommendations

Gopal Sarma, James Koppel, Gregory Malecha, Patrick Schultz, Eric Drexler, Ramana Kumar, Cody Roux, Philip Zucker

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

Comments 6 pages

Controlling Synthetic Characters in Simulations: A Case for Cognitive Architectures and Sigma

Volkan Ustun, Paul S. Rosenbloom, Seyed Sajjadi, Jeremy Nuttal

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

Journal ref Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC) 2018

Adaptive Synthetic Characters for Military Training

Volkan Ustun, Rajay Kumar, Adam Reilly, Seyed Sajjadi, Andrew Miller

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

Journal ref 2020 Interservice/Industry Training, Simulation, and Education Conference (I/ITSEC)

Transformers as Soft Reasoners over Language

Peter Clark, Oyvind Tafjord, Kyle Richardson

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

Comments IJCAI 2020

Knowledge Authoring and Question Answering with KALM

Tiantian Gao

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

Comments In Proceedings ICLP 2019, arXiv:1909.07646

Journal ref EPTCS 306, 2019, pp. 389-395

Bayesian Action Decoder for Deep Multi-Agent Reinforcement Learning

Jakob N. Foerster, Francis Song, Edward Hughes, Neil Burch, Iain Dunning, Shimon Whiteson, Matthew Botvinick, Michael Bowling

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

HOL(y)Hammer: Online ATP Service for HOL Light

Cezary Kaliszyk, Josef Urban

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

Six Llamas: Comparative Religious Ethics Through LoRA-Adapted Language Models

六头 llama：通过 LoRA 调整的语言模型进行比较宗教伦理研究

Chad Coleman, W. Russell Neuman, Manan Shah, Ali Dasdan, Matthew Crispi, Morris Chiang, Zack Leitman, Mustafa Poonawala

机构 * Meta 纠错

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

AI总结 本文通过 LoRA 调整的语言模型比较不同宗教文本对伦理推理的影响，发现调整模型在伦理推理上与基础模型有系统性差异，并在不同温度设置下表现出稳定性与多样性。

Comments 51 pages, 14 figures. We present Six Llamas, a comparative study examining whether Llama-3.1-8B models fine-tuned on distinct religious corpora encode systematically different patterns of ethical reasoning. Five LoRA-adapted variants are constructed for Christianity, Islam, Judaism, Hinduism, and Buddhism. For theoretical background on the condensate comparative method, see arXiv:2603.07329

Investigating Language Model Capabilities to Represent and Process Formal Knowledge: A Preliminary Study to Assist Ontology Engineering

Hanna Abi Akl

机构 * Université Côte d’Azur（法国滨海大学） ； Inria（法国国家信息与自动化技术研究所） ； CNRS（法国国家科学研究中心） ； I3S（国际科学计算研究所） ； Sophia Antipolis, France（法国索菲亚大学） ； Data ScienceTech Institute (DSTI)（数据科学技术研究所） 纠错

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

Comments accepted for the International Joint Conference on Rules and Reasoning (RuleML+RR) 2025

Expected Value Alignment for Generative Reward Modeling in Formal Mathematics Verification

形式数学验证中生成式奖励建模的期望值对齐

Shihao Ji, Haotao Tan, Zihui Song, Mingyu Li

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

AI总结 提出期望值对齐（EVA）方法，通过从模型词元分布中提取连续分数，在保持生成式奖励模型离散输出的同时实现连续评分，用于Lean 4形式验证。

Comments Withdrawn due to serious concerns regarding the authenticity and accuracy of the listed authorship. The identity of one or more listed authors cannot presently be verified, and the author list may not represent distinct contributors. The manuscript is withdrawn pending institutional review

The Emergence of Lab-Driven Alignment Signatures: A Psychometric Framework for Auditing Latent Bias and Compounding Risk in Generative AI

实验室驱动的对齐签名的出现：一种心理测量框架，用于审计生成AI中的潜在偏见和叠加风险

Dusan Bosnjakovic

机构 * AI Researcher（人工智能研究员） 纠错

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

AI总结 本文提出了一种心理测量框架，用于审计生成AI中的潜在偏见和叠加风险，通过分析九个领先模型的实验室信号，揭示了持续行为聚类的成因。

Comments v2: expanded from 9 to 18 behavioral dimensions and from 4 to 6 developer organizations; revised statistical methodology (rank-based inference with effect-size criterion, replacing variance-decomposition approach); model-level results now reported; references corrected throughout