ToMCAT: Theory-of-Mind for Cooperative Agents in Teams via Multiagent Diffusion Policies

TL;DR

This paper introduces ToMCAT, a novel framework combining theory-of-mind reasoning with diffusion models to generate adaptive, team-aware plans for cooperative agents, improving resource efficiency and dynamic coordination in multiagent systems.

Contribution

It presents a new ToM-conditioned trajectory generation method using diffusion models and an online replanning system for dynamic multiagent coordination.

Findings

Replanning reduces resource usage without performance loss.

ToM and recent observations enable better team adaptation.

Dynamic replanning improves coordination in simulated tasks.

Abstract

In this paper we present ToMCAT (Theory-of-Mind for Cooperative Agents in Teams), a new framework for generating ToM-conditioned trajectories. It combines a meta-learning mechanism, that performs ToM reasoning over teammates' underlying goals and future behavior, with a multiagent denoising-diffusion model, that generates plans for an agent and its teammates conditioned on both the agent's goals and its teammates' characteristics, as computed via ToM. We implemented an online planning system that dynamically samples new trajectories (replans) from the diffusion model whenever it detects a divergence between a previously generated plan and the current state of the world. We conducted several experiments using ToMCAT in a simulated cooking domain. Our results highlight the importance of the dynamic replanning mechanism in reducing the usage of resources without sacrificing team performance. We also show that recent observations about the world and teammates' behavior collected by an agent over the course of an episode combined with ToM inferences are crucial to generate team-aware plans for dynamic adaptation to teammates, especially when no prior information is provided about them.