LogiDynamics：揭示LLM推理中归纳、溯因与演绎逻辑推断的动态机制
LogiDynamics: Unraveling the Dynamics of Inductive, Abductive and Deductive Logical Inferences in LLM Reasoning
AI总结：
本文系统研究了LLM中归纳（System 1）与溯因/演绎（System 2）逻辑推断的动态机制，发现System 2在视觉/符号及困难任务中表现更优，而任务格式显著影响两者相对优势，并证明高级System 2策略可大幅扩展推理能力。
摘要：
现代大型语言模型（LLM）采用多种逻辑推断机制进行推理，这使得对这些方法进行策略性优化对于提升其能力至关重要。本文系统地研究了LLM中归纳（System 1）与溯因/演绎（System 2）推断的比较动态。我们利用受控的类比推理环境，改变了模态（文本、视觉、符号）、难度和任务格式（MCQ / 自由文本）。我们的分析表明，System 2流水线总体表现更优，特别是在视觉/符号模态和更难的任务中，而System 1在文本和较简单的问题上具有竞争力。至关重要的是，任务格式显著影响它们的相对优势，System 1有时在自由文本规则执行中优于System 2。这些核心发现可推广到更广泛的上下文学习中。此外，我们证明了假设选择和迭代细化等高级System 2策略可以大幅扩展LLM推理能力。本研究为策略性部署逻辑推断以增强LLM推理提供了基础见解和可操作指南。资源可在https://github.com/HKUST-KnowComp/LogiDynamics获取。
英文摘要：
Modern large language models (LLMs) employ diverse logical inference mechanisms for reasoning, making the strategic optimization of these approaches critical for advancing their capabilities. This paper systematically investigate the comparative dynamics of inductive (System 1) versus abductive/deductive (System 2) inference in LLMs. We utilize a controlled analogical reasoning environment, varying modality (textual, visual, symbolic), difficulty, and task format (MCQ / free-text). Our analysis reveals System 2 pipelines generally excel, particularly in visual/symbolic modalities and harder tasks, while System 1 is competitive for textual and easier problems. Crucially, task format significantly influences their relative advantage, with System 1 sometimes outperforming System 2 in free-text rule-execution. These core findings generalize to broader in-context learning. Furthermore, we demonstrate that advanced System 2 strategies like hypothesis selection and iterative refinement can substantially scale LLM reasoning. This study offers foundational insights and actionable guidelines for strategically deploying logical inference to enhance LLM reasoning. Resources are available at https://github.com/HKUST-KnowComp/LogiDynamics.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
