让大模型学会分辨矛盾信息，自动推理出正确答案。
Disentangling Reasoning Logic to Resolve Explicit Knowledge Conflicts
- 将矛盾信息拆解为独立推理路径，用文本+图结构表示
- 用可验证奖励的强化学习训练模型，提升逻辑一致性
- 70亿参数模型性能超越GPT-4o和GPT-5.1
显式知识冲突指检索到的上下文包含相互矛盾的信息，这给大型语言模型（LLMs）在整合多样化数据源时带来根本性挑战。核心难点在于叙事关系复杂且冲突模式多样，常超出标准骨干架构的推理能力。本文提出 extsc{Kcr}（知识冲突推理）框架，通过系统化梳理矛盾背后的逻辑来裁决冲突。 extsc{Kcr} 将冲突上下文分解为离散的推理轨迹，采用文本与图的混合表示以促进系统化理解。随后，利用带有可验证奖励的强化学习（RLVR）范式，训练模型生成最大化逻辑一致性的推理策略，同时抑制由矛盾证据引出的虚假路径。大量实验表明， extsc{Kcr} 显著提升性能：经 extsc{Kcr} 增强的 7B 模型在复杂任务上表现显著优于 GPT-4o 和 GPT-5.1 等领先专有模型。代码已开源于 https://github.com/zhengxianda/KCR。
原文摘要 · Abstract (English)
Explicit knowledge conflicts, occurring when retrieved contexts contain contradictory information, pose a fundamental challenge for Large Language Models (LLMs) as they integrate increasingly diverse data sources. The core difficulty lies in the complexity of entangled narratives and heterogeneous conflict patterns, which frequently exceeds the reasoning capacity of standard backbone architectures. We propose \textbf{\textsc{Kcr}} (Knowledge Conflict Reasoning), a framework that adjudicates contradictions by systematically structuring their underlying logic. \textsc{Kcr} disentangles conflicting contexts into discrete sets of reasoning traces, utilizing a hybrid representation of text and graphs to facilitate systematic comprehension. It then employs a Reinforcement Learning with Verifiable Rewards (RLVR) paradigm to instill a reasoning policy that maximizes logical consistency while suppressing spurious paths derived from contradictory evidence. Extensive evaluations demonstrate that \textsc{Kcr} yields substantial performance gains. Notably, a 7B model enhanced by \textsc{Kcr} achieves adjudication capabilities that significantly outperform leading proprietary models, including GPT-4o and GPT-5.1, on complex tasks. Code is available at https://github.com/zhengxianda/KCR.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。