ResearchBenchmarksCooperative Multi-Agent Reinforcement Learning on Multi-Agent Particle Environments (MPE)Cooperative Multi-Agent Reinforcement Learning on Multi-Agent Particle Environments (MPE)Follow91IQM of Normalized ReturnsMARS87.8888.6989.590.31May 9, 2026Evaluation resultsMethodMethodLinksIQM of Normalized ReturnsMARS2026.0591MAPPO2026.0589MASPO2026.0588