ResearchBenchmarksOffline Reinforcement Learning on DMControl reacher-easy (expert)Follow58.19Normalized ScoreIQL + S2P25.346833.873442.450.9266Sep 30, 2022Evaluation ResultsMethodMethodLinksNormalized ScoreIQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0958.19CQLS2P Augmentation=falseS2P Augmentation=false2022.0957.68IQLS2P Augmentation=falseS2P Augmentation=false2022.0952.13SLAC-off + S2PS2P Augmentation=trueS2P Augmentation=true2022.0942.85CQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0932.54SLAC-offS2P Augmentation=falseS2P Augmentation=false2022.0926.61