arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新，休息一下吧，周末愉快~~
arXiv 2605.01482cs.AI
收藏

通过群体相对策略优化在结构因果模型中 grounding 多跳推理

Grounding Multi-Hop Reasoning in Structural Causal Models via Group Relative Policy Optimization

School of Computer Science and Technology, Xinjiang University, Urumqi, China（新疆大学计算机科学与技术学院，乌鲁木齐，中国）；Beijing Institute of Technology, Beijing, China（北京理工大学，北京，中国）；Military Science Information Research Center, Academy of Military Science, Beijing, China（军事科学院军事科学信息研究中心，北京，中国）

Yunhan Bu, Quan Zhang, Huaping Zhang, Guotong Geng, Chunxiao Gao, Askar Hamdulla, Juan Wang, Qiuchi Li, Baohua Zhang, Shuai Lei, Yunbo Cao, Zhunchen Luo

阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出基于结构因果模型的多跳事实验证框架，通过群体相对策略优化解决推理链长度与准确率的平衡问题，实验表明其在HoVer和EX-FEVER数据集上优于现有方法。

摘要：

多跳事实验证（MHFV）需要在不同证据间进行复杂推理，这对大语言模型（LLMs）构成重大挑战，因其常出现幻觉和逻辑断裂。现有方法虽然通过思维链（CoT）提高透明度，但缺乏对证据与主张之间因果依赖的显式建模。本文引入一种新颖框架，将推理 grounding 在结构因果模型（SCM）中，将验证视为构造性因果推理过程。我们实证发现推理链长度与准确率之间存在'倒U型'相关性，揭示过度结构复杂性会降低性能。为此，我们提出一种基于规则的强化学习策略，使用群体相对策略优化（GRPO）。该方法动态优化结构深度与简洁性的权衡。在HoVer和EX-FEVER上的大量实验表明，我们的SCM-GRPO框架显著优于现有最先进基线，提供了一种可靠且可解释的复杂事实验证解决方案。

英文摘要：

Multi-Hop Fact Verification requires complex reasoning across disparate evidence, posing significant challenges for Large Language Models , which may suffer from hallucinations and fractured logical chains. Existing methods, while improving transparency via Chain-of-Thought , often lack explicit modeling of the structural dependencies between evidence and claims. In this work, we introduce an SCM-inspired framework that grounds reasoning in explicit directed dependency graphs, treating verification as a constructive structural reasoning process rather than full causal inference with interventions or counterfactual semantics. We empirically identify an "inverted U-shaped" correlation between reasoning-chain length and accuracy, revealing that excessive structural complexity can degrade performance. To address this, we propose a rule-based reinforcement learning strategy using Group Relative Policy Optimization. This approach dynamically optimizes the trade-off between structural depth and conciseness. Extensive experiments on HoVer and EX-FEVER demonstrate that our SCM-GRPO framework outperforms strong baselines while producing more traceable reasoning structures for complex fact verification.

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。

↑