Competition and Cooperation of LLM Agents in Games
This addresses the problem of understanding strategic behavior in competitive multi-agent settings for AI researchers, providing insights into LLM agent dynamics.
The paper studied LLM agent interactions in network resource allocation and Cournot competition games, finding that they tend to cooperate rather than converge to Nash equilibria when given multi-round prompts and non-zero-sum contexts, with fairness reasoning identified as a key factor.
Large language model (LLM) agents are increasingly deployed in competitive multi-agent settings, raising fundamental questions about whether they converge to equilibria and how their strategic behavior can be characterized. In this paper, we study LLM agent interactions in two standard games: a network resource allocation game and a Cournot competition game. Rather than converging to Nash equilibria, we find that LLM agents tend to cooperate when given multi-round prompts and non-zero-sum context. Chain-of-thought analysis reveals that fairness reasoning is central to this behavior. We propose an analytical framework that captures the dynamics of LLM agent reasoning across rounds and explains these experimental findings.