ResearchBenchmarksOffline Reinforcement Learning on Walker2D Medium-Replay Noise 12Follow70.6Normalized ReturnI-TAP35.884844.897453.9162.9226Feb 21, 2026Evaluation ResultsMethodMethodLinksNormalized ReturnI-TAPModel Category=Model-B...Model Category=Model-Based2026.0270.61R2RModel Category=Model-B...Model Category=Model-Based2026.0265.63L-MAPModel Category=Model-B...Model Category=Model-Based2026.0259.16TAPModel Category=Model-B...Model Category=Model-Based2026.0255.15CQLModel Category=Model-FreeModel Category=Model-Free2026.0250.33IQLModel Category=Model-FreeModel Category=Model-Free2026.0245.13DTModel Category=Model-FreeModel Category=Model-Free2026.0237.22