Learning to cooperate: Emergent communication in multi-agent navigation

TL;DR

This paper demonstrates that artificial agents in cooperative navigation tasks develop interpretable, spatially grounded communication protocols with compositional structure, resembling natural language properties, which enhance their task efficiency and interpretability.

Contribution

The study shows that multi-agent systems can develop interpretable, spatially grounded, and compositional communication protocols during cooperative navigation tasks.

Findings

Agents learn spatially clustered, interpretable signals.

Signals refer to specific locations and directions.

Protocols exhibit basic compositional structure.

Abstract

Emergent communication in artificial agents has been studied to understand language evolution, as well as to develop artificial systems that learn to communicate with humans. We show that agents performing a cooperative navigation task in various gridworld environments learn an interpretable communication protocol that enables them to efficiently, and in many cases, optimally, solve the task. An analysis of the agents' policies reveals that emergent signals spatially cluster the state space, with signals referring to specific locations and spatial directions such as "left", "up", or "upper left room". Using populations of agents, we show that the emergent protocol has basic compositional structure, thus exhibiting a core property of natural language.