让多个AI辩论时只讨论分歧点，提升复杂问题回答准确率。
Where Reasoning Diverges: Localized Multi-Agent Debate for Multi-Hop Question Answering
- 将推理过程拆成节点，只在最早冲突处展开辩论
- 在四个数据集上十种模型均表现最佳，最高提升7.2个百分点
- 适合需要精准推理的复杂问答任务
多智能体辩论通常在意见不一致时交换完整推理过程，但实际分歧可能仅涉及少数中间结论。我们提出局部化多智能体辩论（LMAD），将代理推理表示为节点，定位最早冲突点，并仅对相关局部段落进行辩论。受保护的决议机制维护共享确定状态，使后续冲突可处理而不需重新审查已接受步骤。我们在四个多跳问答基准上，使用来自四种模型家族的十种骨干模型评估了LMAD。该方法在所有十种骨干模型中均取得最高的宏平均评分准确率，相较于最强传统基线最高提升7.20个百分点。
原文摘要 · Abstract (English)
Multi-agent debate commonly exchanges complete rationales even when disagreements concern only a few intermediate claims. We introduce Localized Multi-Agent Debate (LMAD), an inference-time protocol that represents agent rationales as nodes, locates their earliest conflict, and restricts debate to the corresponding local segments. Guarded resolution extends a shared committed state so that later conflicts can be addressed without reopening accepted steps. We evaluate LMAD on four multi-hop question-answering benchmarks using ten backbones from four model families. Our method achieves the highest macro-averaged judge accuracy across all ten backbones, outperforming the strongest conventional baseline by up to 7.20 percentage points.
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。