ResearchBenchmarksOffline Reinforcement Learning on D4RL hopper-random v2 (test)Offline Reinforcement Learning on D4RL hopper-random v2 (test)Follow698SpikeSPQR575.881,400.192,224.53,048.81Jan 6, 2024Evaluation resultsMethodMethodLinksSpikePerformanceAccept RatioSPQRβ=100β=1002024.01698—97.6SPQRβ=10β=102024.011,603—92.3SPQRβ=1β=12024.011,844—91.4SPQRβ=0.1β=0.12024.012,182—80.4SPQRβ=0.01β=0.012024.012,360—59SAC-Min2024.012,434—30.4EDACη=1η=12024.013,751—0