[EMNLP 2025 Wordplay] LLM-Hanabi Evaluating Multi-Agent Gameplays with Theory-of-Mind and Rationale Inference in Imperfect Information Collaboration Game - View it on GitHub