ResearchBenchmarksMulti-Agent Reinforcement Learning on SMAC (3m → 1c3s5z)Multi-Agent Reinforcement Learning on SMAC (3m → 1c3s5z)Follow3.8Interactions to 80% Reward (x10^5)GCT-MARL0.29223.97147.6571.329Jun 23, 2026Evaluation resultsMethodMethodLinksInteractions to 80% Reward (x10^5)Win RateGCT-MARL2026.063.8100MAIL2026.066.22100LA-QT2026.0638.762Baseline2026.0653.466CORAL2026.0669.454DANN2026.0678.761CycleGAN2026.0691.565