Developing, Evaluating and Scaling Learning Agents in Multi-Agent Environments

TL;DR

This paper reviews DeepMind's work on multi-agent learning, focusing on complex environments, game theory, and social dilemmas, aiming to advance understanding through benchmarks and open challenges.

Contribution

It provides a taxonomy of multi-agent research challenges and summarizes recent DeepMind developments in complex multi-agent environments.

Findings

DeepMind's multi-agent systems simulate social dilemmas and team coordination.

Development of benchmarks for complex multi-agent tasks.

Identification of open challenges in multi-agent learning.

Abstract

The Game Theory & Multi-Agent team at DeepMind studies several aspects of multi-agent learning ranging from computing approximations to fundamental concepts in game theory to simulating social dilemmas in rich spatial environments and training 3-d humanoids in difficult team coordination tasks. A signature aim of our group is to use the resources and expertise made available to us at DeepMind in deep reinforcement learning to explore multi-agent systems in complex environments and use these benchmarks to advance our understanding. Here, we summarise the recent work of our team and present a taxonomy that we feel highlights many important open challenges in multi-agent research.