ResearchBenchmarksOffline Reinforcement Learning on D4RL Maze Total v2 (test)Follow609.5Normalized ScoreA2PO-21.156142.572306.3470.028Mar 12, 2024Evaluation ResultsMethodMethodLinksNormalized ScoreA2PO2024.03609.5Diffusion-QL2024.03485IQL2024.03307.3EQL2024.03300.8TD3+BC2024.03260.6LAPO2024.03243.4AWAC2024.03220.8CQL2024.03175.8CQL+AW2024.03170.5BCQ2024.03147.5BC2024.0347.9MOPO2024.033.1