ResearchBenchmarksMulti-Agent Reinforcement Learning on SMAC 3m to 3s5zvs3s6zMulti-Agent Reinforcement Learning on SMAC 3m to 3s5zvs3s6zFollow72Win RateASALT53.2858.146367.86Jun 23, 2026Evaluation resultsMethodMethodLinksWin RateASALT2026.0672Baselinetraining=from scratch...training=from scratch using MAPPO2026.0666CycleGANtransfer_scheme=agent-...transfer_scheme=agent-wise transfer2026.0665LA-QTransformer2026.0662DANNtransfer_scheme=agent-...transfer_scheme=agent-wise transfer2026.0661CORALtransfer_scheme=agent-...transfer_scheme=agent-wise transfer2026.0654