Meta-Moderator: Empowering Multi-Agent Debate with Meta-Cognition

2026-08-24

AI Summary

论文提出Meta-Moderator框架，通过元认知过程动态调控多智能体辩论，解决现有方法中冗余讨论和证据聚合不可靠的问题。

Abstract

Multi-agent debate can improve large language model reasoning by eliciting diverse hypotheses and critiques, yet its performance is often constrained by weak moderation. Common pipelines rely on fixed budgets, agreement-based stopping, or untrained judges, leading to redundant deliberation and unreliable evidence aggregation. We cast moderation as a meta-cognitive process, monitoring debate utility, controlling deliberation, and adjudicating a final answer, and introduce Meta-Moderator, a learnable framework that dynamically regulates debate and decides when to finalize an answer. Meta-Moderator is trained independently of the debaters via outcome-driven policy optimization, making debate regulation an explicit capability rather than an incidental effect of prompting. Across five benchmarks, Meta-Moderator outperforms widely used decision layers and transfers across tasks and system configurations. Further analyses show that it allocates debate more selectively and reduces mis-aggregation after informative hypotheses appear.

Problem

• multi-agent debate
• moderation
• debate utility

Innovation

• Meta-Moderator
• meta-cognitive process
• dynamic regulation
• outcome-driven policy optimization
• debate utility