Theory of Mind for Multi-Agent Collaboration via Large Language Models
This work addresses multi-agent collaboration challenges for AI systems, though it is incremental in exploring LLM applications in this domain.
The study evaluated LLM-based agents in a multi-agent cooperative text game with Theory of Mind tasks, finding emergent collaborative behaviors and high-order ToM capabilities, but limitations in planning due to context management and hallucination issues, which were mitigated by explicit belief state representations to enhance performance and ToM inference accuracy.
While Large Language Models (LLMs) have demonstrated impressive accomplishments in both reasoning and planning, their abilities in multi-agent collaborations remains largely unexplored. This study evaluates LLM-based agents in a multi-agent cooperative text game with Theory of Mind (ToM) inference tasks, comparing their performance with Multi-Agent Reinforcement Learning (MARL) and planning-based baselines. We observed evidence of emergent collaborative behaviors and high-order Theory of Mind capabilities among LLM-based agents. Our results reveal limitations in LLM-based agents' planning optimization due to systematic failures in managing long-horizon contexts and hallucination about the task state. We explore the use of explicit belief state representations to mitigate these issues, finding that it enhances task performance and the accuracy of ToM inferences for LLM-based agents.