TeamCraft: A Benchmark for Multi-Modal Multi-Agent Systems in Minecraft

TL;DR

TeamCraft introduces a comprehensive multi-modal multi-agent benchmark in Minecraft to evaluate and improve generalization of collaborative agents across diverse tasks and environments.

Contribution

The paper presents a new benchmark, TeamCraft, with extensive tasks, demonstrations, and evaluation protocols for multi-modal multi-agent collaboration in Minecraft.

Findings

Existing models struggle with generalizing to new goals and scenes.

Significant challenges remain in multi-agent collaboration in complex environments.

The benchmark reveals key limitations of current approaches.

Abstract

Collaboration is a cornerstone of society. In the real world, human teammates make use of multi-sensory data to tackle challenging tasks in ever-changing environments. It is essential for embodied agents collaborating in visually-rich environments replete with dynamic interactions to understand multi-modal observations and task specifications. To evaluate the performance of generalizable multi-modal collaborative agents, we present TeamCraft, a multi-modal multi-agent benchmark built on top of the open-world video game Minecraft. The benchmark features 55,000 task variants specified by multi-modal prompts, procedurally-generated expert demonstrations for imitation learning, and carefully designed protocols to evaluate model generalization capabilities. We also perform extensive analyses to better understand the limitations and strengths of existing approaches. Our results indicate that existing models continue to face significant challenges in generalizing to novel goals, scenes, and unseen numbers of agents. These findings underscore the need for further research in this area. The TeamCraft platform and dataset are publicly available at https://github.com/teamcraft-bench/teamcraft.