ResearchBenchmarksOffline Reinforcement Learning on DMControl walker-walk (mixed)Offline Reinforcement Learning on DMControl walker-walk (mixed)Follow97.84Normalized ScoreCQL + S2P26.267244.848663.4382.0114Sep 30, 2022Evaluation resultsMethodMethodLinksNormalized ScoreCQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0997.84CQLS2P Augmentation=falseS2P Augmentation=false2022.0997.18IQLS2P Augmentation=falseS2P Augmentation=false2022.0996.07IQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0995.49SLAC-off + S2PS2P Augmentation=trueS2P Augmentation=true2022.0992.6SLAC-offS2P Augmentation=falseS2P Augmentation=false2022.0929.02