List your company

Continuous Control on D4RL Hopper medium

103.6Normalized ReturnOFQL
17.2839.6962.184.51Oct 1, 2021May 25, 2022Jan 16, 2023Sep 10, 2023May 3, 2024Dec 25, 2024Aug 19, 2025

Evaluation results

MethodLinks
OFQLCategory=One-step Flow Policies2025.08103.6
DDCategory=Diffusion Planners2025.0898.2
MORELalgorithm class=model-based2021.1095.4
DQLCategory=Multi-step Diffusion Policies2025.0890.5
DiffuserCategory=Diffusion Planners2025.0889.5
FQLCategory=One-step Flow Policies2025.0874.5
CORALTraining Type=Offline2025.0873.15
EDPCategory=Multi-step Diffusion Policies2025.0872.6
ADTraining Type=Offline2025.0872.58
ROMI-CQLbase algorithm=CQL2021.1072.3
ATTraining Type=Offline2025.0870.45
TD3+BCTraining Type=Offline2025.0870.44
IDQLCategory=Multi-step Diffusion Policies2025.0870.1
DTTraining Type=Offline2025.0869.42
CQLalgorithm class=model-free2021.1066.6
IQLCategory=Non-Diffusion Policies2025.0866.3
TD3-BCCategory=Non-Diffusion Policies2025.0859.3
BCCategory=Non-Diffusion Policies2025.0852.9
KIOAmount of data used=5k, Evaluation episodes=1002024.1050.2
Teacher2024.1047.2
BC2021.1040
BC(TD3+BC)Amount of data used=1 million samples, Seeds=102024.1030
BC(CQL)Amount of data used=1 million samples, Seeds=32024.1029
IOAmount of data used=5k2024.1020.6