ResearchBenchmarksMulti-Agent Reinforcement Learning on SMAC (8m → 3m) (Efficiency and Win Rate)Multi-Agent Reinforcement Learning on SMAC (8m → 3m) (Efficiency and Win Rate)Follow80,000Environment Interactions (x10^5)GCT-MARL63,920172,460281,000389,540Jun 23, 2026Evaluation resultsMethodMethodLinksEnvironment Interactions (x10^5)Final Win RateGCT-MARL2026.0680,000100LA-QT2026.06142,00094MAIL2026.06190,000100DANN2026.06280,00092CycleGAN2026.06310,00091CORAL2026.06330,00094Baseline2026.06482,00096