List your company
ResearchFormerly WizWand

Continuous Control on D4RL Walker2d medium

88.86Normalized ReturnCORAL
-3.554420.437844.4368.4222Feb 16, 2021Nov 15, 2021Aug 14, 2022May 14, 2023Feb 10, 2024Nov 8, 2024Aug 8, 2025
Updated 4mo ago

Evaluation results

MethodLinks
CORALTraining Type=Offline2025.0888.86
ATTraining Type=Offline2025.0888.71
TD3+BCTraining Type=Offline2025.0886.91
ADTraining Type=Offline2025.0885.96
COMBO2021.0281.9
BRAC-v2021.0281.1
MOREL2021.0277.8
BRAC-p2021.0277.5
DTTraining Type=Offline2025.0874.7
KIOAmount of data used=5k, Evaluation episodes=1002024.1074.6
CQL2021.0274.5
Teacher2024.1068.1
BEAR2021.0259.1
MOPO2021.0217.8
BC(TD3+BC)Amount of data used=1 million samples, Seeds=102024.1011.4
BC(CQL)Amount of data used=1 million samples, Seeds=32024.106.6
BC2021.026.6
SAC-off2021.020.9
IOAmount of data used=5k2024.100