ResearchBenchmarksMulti-Agent Reinforcement Learning on SMAC 25mFollow100Win RateMAPPO86.37689.91393.4596.987Jan 22, 2026Evaluation ResultsMethodMethodLinksWin RateStd DevMAPPODifficulty=Hard, Sight...Difficulty=Hard, Sight Range=92026.011002.7DG-MAPPODifficulty=Hard, Sight...Difficulty=Hard, Sight Range=92026.0195.33.1HAPPODifficulty=Hard, Sight...Difficulty=Hard, Sight Range=92026.01952MAT-DecDifficulty=Hard, Sight...Difficulty=Hard, Sight Range=92026.0186.95.6