arXiv 2605.268932026-05-26cs.CLcs.AI

GeoFaith: 时空双视角下的忠实思维链

GeoFaith: A Spatio-Temporal Dual View of Faithful Chain-of-Thought

Xidian University（西安电子科技大学）；Xi’an Jiaotong University（西安交通大学）；Mohamed bin Zayed University of Artificial Intelligence（穆罕默德·本·扎耶德人工智能大学）；The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州））；University of Science and Technology of China（中国科学技术大学）

Weijiang Lv, Wentong Zhao, Jiayu Wang, Yuhao Wu, Jiaheng Wei, Xiaobo Xia

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

针对思维链推理中的事后合理化问题，提出基于潜在几何结构和熵动力学的时空框架GeoFaith，通过可扩展的引导流水线构建忠实性检测器并联合优化结果正确性、过程忠实性和轨迹一致性。

摘要：

思维链推理推动了大型语言模型的发展，但基于结果的监督导致了普遍的事后合理化，产生了看似合理但不忠实的推理链。大多数先前的忠实性评估方法要么不可扩展、昂贵，要么不可靠。我们提出GeoFaith，一个利用潜在几何结构和熵动力学来诊断和强制忠实推理的时空框架。我们开发了一个可扩展的引导流水线，将四个领域的步骤级标注从1k扩展到20k样本，训练了一个在标准基准上优于GPT-5的8B忠实性检测器，并设计了一个忠实性感知的强化学习框架，联合优化结果正确性、过程忠实性和轨迹一致性。实验表明，所提出的方法在忠实性检测和下游推理上均取得了优越性能，生成了更短、更可解释的链，且不牺牲准确性。我们的代码将公开提供。

英文摘要：

Chain-of-Thought (CoT) reasoning has advanced large language models (LLMs), but outcome-based supervision leads to pervasive post-hoc rationalization, producing plausible yet unfaithful reasoning chains. Most prior faithfulness assessment methods are either unscalable, expensive, or unreliable. We propose GeoFaith, a spatio-temporal framework that leverages latent geometric structure and entropy dynamics to diagnose and enforce faithful reasoning. We develop a scalable bootstrapping pipeline expanding step-level annotations from 1k to 20k samples across four domains, train an 8B faithfulness detector outperforming GPT-5 on standard benchmarks, and design a faithfulness-aware reinforcement learning framework jointly optimizing outcome correctness, process faithfulness, and trajectory consistency. Experiments show the proposed method achieves superior performance on both faithfulness detection and downstream reasoning, producing shorter, more interpretable chains without sacrificing accuracy. Our code will be made available publicly.

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。