Abstract: This paper presents an exploration of large language models (LLMs) in the context of competitive fighting game AI, addressing the limitations of reinforcement learning (RL) agents that often ...