Theory of Mind for Multi-Agent Collaboration via Large Language Models

Abstract

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.

Accepted to EMNLP 2023 (Main Conference). Code available at https://github.com/romanlee6/multi_LLM_comm

References in corpus (17)

• PaLM: Scaling Language Modeling with Pathways
• Sparks of Artificial General Intelligence: Early experiments with GPT-4
• LaMDA: Language Models for Dialog Applications
• Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
• Learning a Generic Value-Selection Heuristic Inside a Constraint Programming Solver
• Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena
• Voyager: An Open-Ended Embodied Agent with Large Language Models
• Language Models as Zero-Shot Planners: Extracting Actionable Knowledge for Embodied Agents
• Dissociating language and thought in large language models
• LLM+P: Empowering Large Language Models with Optimal Planning Proficiency
• Large Language Models Fail on Trivial Alterations to Theory-of-Mind Tasks
• Machine Psychology
• Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents
• Translating Natural Language to Planning Goals with Large-Language Models
• Boosting Theory-of-Mind Performance in Large Language Models via Prompting
• Emergence of Theory of Mind Collaboration in Multiagent Systems
• Theory of Mind as Intrinsic Motivation for Multi-Agent Reinforcement Learning

Cited by in corpus (2)

• From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users
• Evaluating Theory of Mind and Internal Beliefs in LLM-Based Multi-Agent Systems