DynaDebate: Breaking Homogeneity in Multi-Agent Debate with Dynamic Path Generation

TL;DR

DynaDebate introduces a dynamic multi-agent debate framework that generates diverse reasoning paths and emphasizes process critique, significantly improving collaborative decision-making in large language model-based systems.

Contribution

It proposes a novel approach with dynamic path generation, process-centric debate, and trigger-based verification to overcome homogeneity and enhance multi-agent reasoning.

Findings

Outperforms existing MAD methods on multiple benchmarks.

Achieves more diverse and logical reasoning paths.

Reduces errors caused by homogeneous reasoning.

Abstract

Recent years have witnessed the rapid development of Large Language Model-based Multi-Agent Systems (MAS), which excel at collaborative decision-making and complex problem-solving. Recently, researchers have further investigated Multi-Agent Debate (MAD) frameworks, which enhance the reasoning and collaboration capabilities of MAS through information exchange and debate among multiple agents. However, existing approaches often rely on unguided initialization, causing agents to adopt identical reasoning paths that lead to the same errors. As a result, effective debate among agents is hindered, and the final outcome frequently degenerates into simple majority voting. To solve the above problem, in this paper, we introduce Dynamic Multi-Agent Debate (DynaDebate), which enhances the effectiveness of multi-agent debate through three key mechanisms: (1) Dynamic Path Generation and Allocation, which employs a dedicated Path Generation Agent to generate diverse and logical solution paths with adaptive redundancy; (2) Process-Centric Debate, which shifts the focus from surface-level outcome voting to rigorous step-by-step logic critique to ensure process correctness; (3) A Trigger-Based Verification Agent, which is activated upon disagreement and uses external tools to objectively resolve deadlocks. Extensive experiments demonstrate that DynaDebate achieves superior performance across various benchmarks, surpassing existing state-of-the-art MAD methods.