ResearchBenchmarksOffline Reinforcement Learning on DMControl cartpole-swingup (mixed)Follow16.36Normalized ScoreSLAC-off + S2P-5.50080.17465.8511.5254Sep 30, 2022Evaluation ResultsMethodMethodLinksNormalized ScoreSLAC-off + S2PS2P Augmentation=trueS2P Augmentation=true2022.0916.36CQLS2P Augmentation=falseS2P Augmentation=false2022.0914.76SLAC-offS2P Augmentation=falseS2P Augmentation=false2022.0914.51IQLS2P Augmentation=falseS2P Augmentation=false2022.0914.49IQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.0914.04CQL + S2PS2P Augmentation=trueS2P Augmentation=true2022.09-4.66