ResearchBenchmarksOffline Reinforcement Learning on D4RL ant expert v3Follow120Normalized ScoreDROCO55.10471.95288.8105.648Dec 2, 2025Evaluation ResultsMethodMethodLinksNormalized ScoreDROCOShift Type=KinematicShift Type=Kinematic2025.12120IGDFShift Type=KinematicShift Type=Kinematic2025.12119.2DARAShift Type=KinematicShift Type=Kinematic2025.12115.1OTDFShift Type=KinematicShift Type=Kinematic2025.12111.6IQLShift Type=KinematicShift Type=Kinematic2025.12111CQLShift Type=KinematicShift Type=Kinematic2025.1294.3BOSAShift Type=KinematicShift Type=Kinematic2025.1257.6