Emergent Resource Exchange and Tolerated Theft Behavior using Multi-Agent Reinforcement Learning

TL;DR

This paper demonstrates how multi-agent reinforcement learning can lead to the emergence of resource exchange protocols and tolerated theft behaviors in a simulated foraging environment, highlighting novel cooperation mechanisms.

Contribution

Introduces a new resource exchange protocol enabled by campfire-induced congregation, showing emergent cooperation and tolerated theft without explicit punishment mechanisms.

Findings

Agents learn to avoid being cheated in exchanges

Emergence of tolerated theft behavior without punishment

Resource exchange improves cooperation efficiency

Abstract

For decades, the evolution of cooperation has piqued the interest of numerous academic disciplines such as game theory, economics, biology, and computer science. In this work, we demonstrate the emergence of a novel and effective resource exchange protocol formed by dropping and picking up resources in a foraging environment. This form of cooperation is made possible by the introduction of a campfire, which adds an extended period of congregation and downtime for agents to explore otherwise unlikely interactions. We find that the agents learn to avoid getting cheated by their exchange partners, but not always from a third party. We also observe the emergence of behavior analogous to tolerated theft, despite the lack of any punishment, combat, or larceny mechanism in the environment.