Navigating Unreliable Parametric and Contextual Knowledge: Explicit Knowledge Conflict Resolution for LLM Inference

📄 arXiv: 2606.20245v1 📥 PDF

作者: Huang Peng, Jiuyang Tang, Weixin Zeng, Hao Xu, Xiang Zhao

分类: cs.AI

发布日期: 2026-06-18

备注: 12 pages, 3 figures

💡 一句话要点

提出MACR框架以解决LLM推理中的知识冲突问题

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 大型语言模型 知识冲突 多智能体推理 自然语言处理 智能问答系统

📋 核心要点

  1. 现有方法假设模型或上下文可靠，忽视了两者可能同时存在错误的问题，导致知识冲突无法有效解决。
  2. 本文提出MACR框架，通过多智能体推理机制，主动识别和解决知识冲突，提升模型的推理能力。
  3. 实验结果表明，MACR在多个基准测试中显著超越了现有方法，提供了更为准确和可解释的冲突解决方案。

📝 摘要（中文）

大型语言模型（LLMs）在多种语言任务中表现出色，但在整合外部知识时可能会引发内部参数知识与外部信息之间的冲突。现有方法通常假设模型或上下文可靠，忽视了两者可能都存在错误的问题。为了解决这些局限性，本文提出了一种新颖的MACR框架，采用多智能体推理方法，主动解决知识冲突。该框架通过自适应知识评估与检索，量化模型对答案的信心，并引入三个专门的智能体来分析和解决冲突。实验证明，MACR在多个基准测试中显著优于现有最先进的方法，并提供了可解释的冲突解决方案。

🔬 方法详解

问题定义：本文旨在解决大型语言模型在推理过程中因整合外部知识而引发的知识冲突问题。现有方法通常假设模型或上下文是可靠的，未能有效处理两者可能同时存在错误的情况。

核心思路：提出MACR框架，采用多智能体推理机制，主动识别和解决知识冲突。通过自适应知识评估，量化模型对答案的信心，从而决定是外部化内部知识还是检索外部知识。

技术框架：MACR框架包括两个主要模块：自适应知识评估与检索模块，以及多智能体推理模块。前者负责评估模型的信心并生成基本上下文，后者则通过三个专门的智能体进行冲突分析与解决。

关键创新：MACR的核心创新在于引入了多智能体推理机制，能够同时处理多个知识源之间的冲突，而不是简单地优先考虑某一来源。这一设计使得模型在面对不确定性时更具鲁棒性。

关键设计：在自适应知识评估中，采用了修改的语义熵度量来量化模型的信心。此外，多智能体推理框架中的三个智能体分别负责生成显式规则、分析潜在冲突和解决不一致性，确保了冲突解决的全面性与有效性。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果显示，MACR在多个基准测试中显著超越了现有最先进的方法，提升幅度达到XX%（具体数据待补充），并提供了可解释的冲突解决方案，增强了模型的透明性和可信度。

🎯 应用场景

该研究的潜在应用领域包括自然语言处理、智能问答系统和对话系统等。通过有效解决知识冲突，MACR框架能够提升模型在复杂场景下的推理能力，具有广泛的实际价值和未来影响。

📄 摘要（原文）

Large language models (LLMs) have achieved strong performance across a wide range of language-based tasks by leveraging both extensive parametric knowledge and in-context learning ability, enabling them to incorporate external information provided in the input prompt. However, the integration of external knowledge can introduce conflicts, not only between the model's internal parametric knowledge and the external information, but also among multiple pieces of external contexts. Existing approaches typically assume that either the model or the provided context is reliable, overlooking the possibility that both sources may contain errors, and avoid conflicts by privileging one source over the other, rather than actively resolving inconsistencies. To address these limitations, we propose a novel framework MACR for LLM knowledge conflict resolution that moves beyond the conventional binary choice paradigm and incorporates an explicit conflict-resolution mechanism based on a multi-agent reasoning approach. Specifically, we first propose an adaptive knowledge assessment and retrieval approach that employs a modified semantic entropy measure to quantify an LLM's confidence in its answer to a given query. Based on this confidence estimation, MACR either externalizes the model's internal knowledge as textual representations or retrieves relevant external knowledge when internal knowledge is insufficient, generating basic contexts for subsequent reasoning. Then we introduce an inductive multi-agent reasoning framework with three specialized agents that, respectively, induce explicit rules, analyze potential conflicts, and resolve inconsistencies across all available contexts. Empirical results demonstrate that MACR significantly outperforms state-of-the-art baselines across benchmarks, while also providing interpretable resolutions of explicit conflicts.