ResearchBenchmarksConstrained Offline Reinforcement Learning on DSRL PointGoal1Constrained Offline Reinforcement Learning on DSRL PointGoal1Follow0.74Normalized ReturnMPDiffuser0.41760.50130.5850.6687Dec 9, 2025Evaluation resultsMethodMethodLinksNormalized ReturnCostMPDiffusernum_samples=16num_samples=162025.120.740.88BCQ-Lag2025.120.710.98CDT2025.120.691.12BC-All2025.120.650.95CPQ2025.120.570.35COptiDICE2025.120.491.66BC-Safe2025.120.430.54