List your company
ResearchFormerly WizWand

Reinforcement Learning on D4RL Hopper-Medium-Replay

98.33Normalized ReturnCORAL
24.146843.405962.66581.9241Aug 8, 2025Sep 22, 2025Nov 6, 2025Dec 21, 2025Feb 4, 2026Mar 21, 2026May 6, 2026
Updated 4mo ago

Evaluation results

MethodLinks
CORALTraining Type=Offline2025.0898.33
TD3+BCTraining Type=Offline2025.0898.12
ATTraining Type=Offline2025.0896.85
ADTraining Type=Offline2025.0891.32
DTTraining Type=Offline2025.0888.74
Adaptive Policy Selection and Fine-TuningPhase=Online, Online Interaction Budget=160 K2026.0583.6
BestPhase=Offline, Online Interaction Budget=160 K2026.0572.6
OEPhase=Offline, Online Interaction Budget=160 K2026.0572.1
FTPhase=Online, Online Interaction Budget=160 K2026.0535.2
OPEPhase=Offline, Online Interaction Budget=160 K2026.0527