courtroom 风格多智能体辩论与渐进式 RAG 和角色切换用于争议性主张验证
Courtroom-Style Multi-Agent Debate with Progressive RAG and Role-Switching for Controversial Claim Verification
AI总结：
本文提出 PROClaim 框架，通过结构化对抗辩论和角色切换提升争议性主张验证的可靠性，实验显示其在 Check-COVID 数据集上准确率达 81.7%。
摘要：
大型语言模型 (LLMs) 由于幻觉和浅层推理在高风险主张验证中仍不可靠。虽然检索增强生成 (RAG) 和多智能体辩论 (MAD) 有所改进，但受限于单次检索和无结构辩论动态。我们提出 courtroom 风格多智能体框架 PROClaim，将验证重新公式化为结构化、对抗性辩论。我们的方法整合了专门角色 (例如原告、辩护方、法官) 与渐进式 RAG (P-RAG) 来动态扩展和细化证据池。此外，我们采用证据谈判、自我反思和异质多法官聚合来强制校准、鲁棒性和多样性。在零样本评估中，PROClaim 在 Check-COVID 数据集上达到 81.7% 准确率，优于标准多智能体辩论 10.0 个百分点，其中 P-RAG 推动了主要性能提升 (+7.5 pp)。我们最终证明，结构化辩论和模型异质性有效缓解系统性偏差，为可靠的主张验证提供稳健基础。我们的代码和数据在 https://github.com/mnc13/PROClaim 上公开。
英文摘要：
Large language models (LLMs) remain unreliable for high-stakes claim verification due to hallucinations and shallow reasoning. While retrieval-augmented generation (RAG) and multi-agent debate (MAD) address this, they are limited by one-pass retrieval and unstructured debate dynamics. We propose a courtroom-style multi-agent framework, PROClaim, that reformulates verification as a structured, adversarial deliberation. Our approach integrates specialized roles (e.g., Plaintiff, Defense, Judge) with Progressive RAG (P-RAG) to dynamically expand and refine the evidence pool during the debate. Furthermore, we employ evidence negotiation, self-reflection, and heterogeneous multi-judge aggregation to enforce calibration, robustness, and diversity. In zero-shot evaluations on the Check-COVID benchmark, PROClaim achieves 81.7% accuracy, outperforming standard multi-agent debate by 10.0 percentage points, with P-RAG driving the primary performance gains (+7.5 pp). We ultimately demonstrate that structural deliberation and model heterogeneity effectively mitigate systematic biases, providing a robust foundation for reliable claim verification. Our code and data are publicly available at https://github.com/mnc13/PROClaim.
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。
