ResearchBenchmarksOffline Reinforcement Learning on D4RL ant medium-expert v3Follow119Normalized ScoreDROCO99.864104.832109.8114.768Dec 2, 2025Evaluation ResultsMethodMethodLinksNormalized ScoreDROCOShift Type=KinematicShift Type=Kinematic2025.12119IGDFShift Type=KinematicShift Type=Kinematic2025.12112.8IQLShift Type=KinematicShift Type=Kinematic2025.12106.1OTDFShift Type=KinematicShift Type=Kinematic2025.12105.1DARAShift Type=KinematicShift Type=Kinematic2025.12104.8BOSAShift Type=KinematicShift Type=Kinematic2025.12102.5CQLShift Type=KinematicShift Type=Kinematic2025.12100.6