Learning Multi-Agent Communication with Contrastive Learning

TL;DR

This paper introduces a contrastive learning approach for multi-agent communication, treating messages as environment state views, leading to improved coordination, faster learning, and more symmetric, informative communication.

Contribution

It proposes a novel contrastive learning method that maximizes mutual information between messages, enhancing communication effectiveness in multi-agent reinforcement learning.

Findings

Outperforms previous methods in performance and learning speed

Induces more symmetric and global state-aware communication

Effectively captures environment information through learned messages

Abstract

Communication is a powerful tool for coordination in multi-agent RL. But inducing an effective, common language is a difficult challenge, particularly in the decentralized setting. In this work, we introduce an alternative perspective where communicative messages sent between agents are considered as different incomplete views of the environment state. By examining the relationship between messages sent and received, we propose to learn to communicate using contrastive learning to maximize the mutual information between messages of a given trajectory. In communication-essential environments, our method outperforms previous work in both performance and learning speed. Using qualitative metrics and representation probing, we show that our method induces more symmetric communication and captures global state information from the environment. Overall, we show the power of contrastive learning and the importance of leveraging messages as encodings for effective communication.