The Neural MMO Platform for Massively Multiagent Research

TL;DR

Neural MMO is an open-source platform that enables research on large populations of agents in complex, open-ended environments, facilitating studies on multiagent interactions, learning, and cooperation.

Contribution

It introduces Neural MMO as the first platform combining large agent populations, long horizons, open-ended tasks, and modular systems for multiagent research.

Findings

Agents trained in large populations explore more and develop skills over time.

Neural MMO supports complex multi-team cooperation research.

The platform is open source with tools for training, logging, and visualization.

Abstract

Neural MMO is a computationally accessible research platform that combines large agent populations, long time horizons, open-ended tasks, and modular game systems. Existing environments feature subsets of these properties, but Neural MMO is the first to combine them all. We present Neural MMO as free and open source software with active support, ongoing development, documentation, and additional training, logging, and visualization tools to help users adapt to this new setting. Initial baselines on the platform demonstrate that agents trained in large populations explore more and learn a progression of skills. We raise other more difficult problems such as many-team cooperation as open research questions which Neural MMO is well-suited to answer. Finally, we discuss current limitations of the platform, potential mitigations, and plans for continued development.