ResearchBenchmarksMulti-agent Offline Reinforcement Learning on MPE World (Medium-replay)Multi-agent Offline Reinforcement Learning on MPE World (Medium-replay)Follow73.4ScoreCFCQL6.3223.73541.1558.565Sep 22, 2023Evaluation resultsMethodMethodLinksScoreAverage Normalized ScoreCFCQL2023.0973.4—OMAR2023.0942.9—IQL2023.0941.5—ICQL2023.0929.6—MATD3-BC2023.0917.4—MACQL2023.0913.2—MAICQ2023.0912—AWAC2023.098.9—CFCQL2026.03—73.4DoF2026.03—58.6IQL2026.03—41.5MA-TD3+BC2026.03—17.4MACQL2026.03—13.2OMAR2026.03—42.9PLCQL2026.03—107.5SPaCQL2026.03—105.2