ResearchBenchmarksOffline Reinforcement Learning on D4RL Halfcheetah random v0Follow4,114ReturnMOPO-207.2914.652,036.53,158.35Feb 22, 2021Evaluation ResultsMethodMethodLinksReturnMOPOvariant=bootstrapvariant=bootstrap2021.024,114MBPO2021.023,527SAC2021.023,502GELATOdistance=metricdistance=metric2021.022,560GELATOdistance=l2distance=l22021.02512Imitation2021.02-41