ResearchBenchmarksOffline Reinforcement Learning on D4RL Adroit door human cloned expertFollow38.5Normalized ScoreMoMa QL34.54835.57436.637.626May 27, 2026Evaluation ResultsMethodMethodLinksNormalized ScoreMoMa QL2026.0538.5IQL2026.0537.7IFQL2026.0537.7IDQL2026.0537SRPO2026.0536FBRAC2026.0536BC2026.0535.7ReBRAC2026.0535.3FQL2026.0535.3FAWAC2026.0535CAC2026.0534.7