SWE-Debate: Competitive Multi-Agent Debate for Software Issue Resolution

TL;DR

SWE-Debate introduces a multi-agent debate framework that enhances software issue resolution by fostering diverse reasoning, leading to more accurate fault localization and patch generation, outperforming existing methods on benchmark tests.

Contribution

It presents a novel multi-agent debate approach that improves fault localization and issue resolution in software engineering tasks, surpassing prior independent exploration methods.

Findings

Achieves state-of-the-art results on SWE-bench benchmark.

Outperforms baseline methods significantly.

Enhances fault localization accuracy through debate mechanism.

Abstract

Issue resolution has made remarkable progress thanks to the advanced reasoning capabilities of large language models (LLMs). Recently, agent-based frameworks such as SWE-agent have further advanced this progress by enabling autonomous, tool-using agents to tackle complex software engineering tasks. While existing agent-based issue resolution approaches are primarily based on agents' independent explorations, they often get stuck in local solutions and fail to identify issue patterns that span across different parts of the codebase. To address this limitation, we propose SWE-Debate, a competitive multi-agent debate framework that encourages diverse reasoning paths and achieves more consolidated issue localization. SWE-Debate first creates multiple fault propagation traces as localization proposals by traversing a code dependency graph. Then, it organizes a three-round debate among specialized agents, each embodying distinct reasoning perspectives along the fault propagation trace. This structured competition enables agents to collaboratively converge on a consolidated fix plan. Finally, this consolidated fix plan is integrated into an MCTS-based code modification agent for patch generation. Experiments on the SWE-bench benchmark show that SWE-Debate achieves new state-of-the-art results in open-source agent frameworks and outperforms baselines by a large margin.