AI RESEARCH
Stay Focused: Problem Drift in Multi-Agent Debate
arXiv CS.CL
•
ArXi:2502.19559v3 Announce Type: replace Multi-agent debate - multiple instances of large language models discussing problems in turn-based interaction - has shown promise for solving knowledge and reasoning tasks. However, these methods show limitations when solving complex problems that require longer reasoning chains. We analyze how multi-agent debate drifts away from the initial problem over multiple turns, thus harming task performance.