ResearchBenchmarksCooperative Multi-Agent Reinforcement Learning on SMAXCooperative Multi-Agent Reinforcement Learning on SMAXFollow0.92IQM of Normalized ReturnsMARS0.8160.8430.870.897May 9, 2026Evaluation resultsMethodMethodLinksIQM of Normalized ReturnsMARS2026.050.92MAPPO2026.050.89MASPO2026.050.82