arXiv 2601.060022026-01-13cs.CLcs.AI

思维的分子结构：映射长链推理的拓扑结构

The Molecular Structure of Thought: Mapping the Topology of Long Chain-of-Thought Reasoning

LARG, SCIR, Harbin Institute of Technology（LARG、SCIR、哈尔滨工业大学）；Peking University（北京大学）；Nanjing University（南京大学）；Central South University（中南大学）

Qiguang Chen, Yantao Du, Ziniu Li, Jinhao Liu, Songyao Duan, Jiarui Guo, Minghao Liu, Jiaheng Liu, Tong Yang, Ge Zhang, Libo Qin, Wanxiang Che, Wenhao Huang

备注：: Preprint

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出通过分子结构类比理解长链推理的拓扑特性，引入有效语义异构体概念，并提出Mole-Syn方法提升长链推理性能和稳定性。

摘要：

大型语言模型（LLMs）往往难以从人类或非长链推理（Long CoT）LLM的模仿中学习有效的长链推理。为了理解这一点，我们提出有效的和可学习的长链推理轨迹特征在统一视图中具有稳定的分子-like结构，这些结构由三种相互作用类型形成：深度推理（共价-like）、自我反思（氢键-like）和自我探索（范德华-like）。对压缩轨迹的分析表明，这些结构源于长链推理微调，而不是关键词模仿。我们引入了有效语义异构体，并表明只有促进快速熵收敛的键才支持稳定的长链推理学习，而结构竞争会损害训练。基于这些发现，我们提出了Mole-Syn，一种分布转移图方法，用于引导有效长链推理结构的合成，在多个基准上提升了性能和强化学习的稳定性。

英文摘要：

Large language models (LLMs) often fail to learn effective long chain-of-thought (Long CoT) reasoning from human or non-Long-CoT LLMs imitation. To understand this, we propose that effective and learnable Long CoT trajectories feature stable molecular-like structures in unified view, which are formed by three interaction types: Deep-Reasoning (covalent-like), Self-Reflection (hydrogen-bond-like), and Self-Exploration (van der Waals-like). Analysis of distilled trajectories reveals these structures emerge from Long CoT fine-tuning, not keyword imitation. We introduce Effective Semantic Isomers and show that only bonds promoting fast entropy convergence support stable Long CoT learning, while structural competition impairs training. Drawing on these findings, we present Mole-Syn, a distribution-transfer-graph method that guides synthesis of effective Long CoT structures, boosting performance and RL stability across benchmarks.

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。