L-MAD: A Systematic Evaluation of Multi-Agent Debate Structures in Legal Reasoning
L-MAD: A Systematic Evaluation of Multi-Agent Debate Structures in Legal Reasoning
L-MAD：法律推理中多智能体辩论结构的系统性评估
Abstract: While multi-agent debate (MAD) frameworks have shown significant potential in general reasoning, their effectiveness in highly structured, knowledge-heavy legal domains remains under-explored. 摘要： 尽管多智能体辩论（MAD）框架在通用推理方面展现出了巨大潜力，但它们在高度结构化、知识密集型的法律领域中的有效性仍未得到充分探索。
In this work, we introduce the Legal Multi-Agent Debate (L-MAD) framework to systematically evaluate different debate structures and aggregation methods within Legal Textual Entailment. 在这项工作中，我们引入了法律多智能体辩论（L-MAD）框架，旨在系统地评估法律文本蕴含任务中不同的辩论结构和聚合方法。
By assigning distinct expert personas to multiple agents, L-MAD improves upon strong single-agent baselines by up to 8%. 通过为多个智能体分配不同的专家角色，L-MAD 在强大的单智能体基准模型基础上，性能提升了高达 8%。
Furthermore, analyzing how debate scales reveals a clear trade-off: increasing the agent population reduces inconsistency and improves accuracy, whereas extending discussion rounds induces a detrimental \textit{over-deliberation drift} where agents reinforce each other’s mistakes. 此外，对辩论规模扩展的分析揭示了一个明显的权衡：增加智能体数量可以减少不一致性并提高准确性，而延长讨论轮次则会导致有害的“过度审议漂移”（over-deliberation drift），即智能体会相互强化彼此的错误。
Ultimately, our findings outline the practical boundaries and safety margins of deploying collaborative multi-agent systems in high-stakes legal reasoning environments. 最终，我们的研究结果勾勒出了在涉及高风险的法律推理环境中部署协作式多智能体系统的实际边界与安全余量。
Paper Details:
- Authors: Tan-Minh Nguyen, Hoang-Trung Nguyen, Huu-Dong Nguyen, Dinh-Truong Do, Thi-Hai-Yen Vuong, Le-Minh Nguyen
- Submission Date: 10 Jul 2026
- arXiv ID: 2607.09099
论文详情：
- 作者： Tan-Minh Nguyen, Hoang-Trung Nguyen, Huu-Dong Nguyen, Dinh-Truong Do, Thi-Hai-Yen Vuong, Le-Minh Nguyen
- 提交日期： 2026年7月10日
- arXiv ID： 2607.09099