ResearchBenchmarksOffline Reinforcement Learning on DMControl finger-spin (expert)Follow98.42Normalized ScoreIQL21.803241.694161.58581.4759Sep 30, 2022Evaluation ResultsMethodMethodLinksNormalized ScoreIQLS2P Augmentation=falseS2P Augmentation=false2022.0998.42CQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0997.25IQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0994.42SLAC-off + S2PS2P Augmentation=trueS2P Augmentation=true2022.0981.05CQLS2P Augmentation=falseS2P Augmentation=false2022.0973.07SLAC-offS2P Augmentation=falseS2P Augmentation=false2022.0924.75