Our work on Game Reasoning Arena is out. See LLM agents compete, strategize and adapt in dynamic game environments.