Emergent Coordination Through Competition
This addresses the challenge of large-scale multi-agent training in continuous control for AI research, but it is incremental as it builds on existing methods like population-based training.
The paper tackled the problem of emergent cooperative behaviors in reinforcement learning agents by introducing a competitive multi-agent soccer environment with continuous physics, demonstrating that decentralized, population-based training can lead to agents progressing from random actions to evidence of cooperation.
We study the emergence of cooperative behaviors in reinforcement learning agents by introducing a challenging competitive multi-agent soccer environment with continuous simulated physics. We demonstrate that decentralized, population-based training with co-play can lead to a progression in agents' behaviors: from random, to simple ball chasing, and finally showing evidence of cooperation. Our study highlights several of the challenges encountered in large scale multi-agent training in continuous control. In particular, we demonstrate that the automatic optimization of simple shaping rewards, not themselves conducive to co-operative behavior, can lead to long-horizon team behavior. We further apply an evaluation scheme, grounded by game theoretic principals, that can assess agent performance in the absence of pre-defined evaluation tasks or human baselines.