ResearchBenchmarksMulti-Agent Reinforcement Learning on SMAC MMM2Follow98.9Win RateDG-MAPPO48.9861.9474.987.86Jan 22, 2026Feb 2, 2026Feb 14, 2026Feb 25, 2026Mar 9, 2026Mar 20, 2026Apr 1, 2026Evaluation ResultsMethodMethodLinksWin RateStd DevDG-MAPPODifficulty=Hard+, Sigh...Difficulty=Hard+, Sight Range=92026.0198.91.2GRASPbase_method=MAPPObase_method=MAPPO2026.0496.7—MAT-DecDifficulty=Hard+, Sigh...Difficulty=Hard+, Sight Range=92026.0191.25.3HAPPODifficulty=Hard+, Sigh...Difficulty=Hard+, Sight Range=92026.0188.82MA2E2026.0483.9—MAPPODifficulty=Hard+, Sigh...Difficulty=Hard+, Sight Range=92026.0181.810.1MAPPO2026.0471.7—HAPPO2026.0450.9—