List your company
ResearchFormerly WizWand

Offline Reinforcement Learning on D4RL Walker2d Full-Replay

107.8Normalized ScorePSPO
83.8890.0996.3102.51May 8, 2026
Updated 5mo ago

Evaluation results

MethodLinks
PSPOAveraged across=4 random seeds2026.05107.8
EPQ2026.05107.4
ADM2026.0599.9
DMG2026.0597.5
PMDB2026.0595.4
CQL2026.0594.2
RAMBO2026.0588.3
MOReL2026.0584.8