ResearchBenchmarksOffline Reinforcement Learning on D4RL Ant Medium-Replay v3Follow58.2Normalized ScoreIGDF20.55230.32640.149.874Dec 2, 2025Evaluation ResultsMethodMethodLinksNormalized ScoreIGDFShift Type=KinematicShift Type=Kinematic2025.1258.2IQLShift Type=KinematicShift Type=Kinematic2025.1246.8DROCOShift Type=KinematicShift Type=Kinematic2025.1244.8DARAShift Type=KinematicShift Type=Kinematic2025.1242.1OTDFShift Type=KinematicShift Type=Kinematic2025.1239.6CQLShift Type=KinematicShift Type=Kinematic2025.1239.4BOSAShift Type=KinematicShift Type=Kinematic2025.1222