List your company
ResearchFormerly WizWand

Offline Reinforcement Learning on D4RL Maze maze2d-medium v2 (test)

11,490Normalized ScoreA2PO
-386.82,696.65,7808,863.4Mar 12, 2024
Updated 8mo ago

Evaluation results

MethodLinks
A2PO2024.0311,490
Diffusion-QL2024.0310,060
LAPO2024.034,320
EQL2024.033,630
TD3+BC2024.033,350
AWAC2024.033,140
IQL2024.032,570
CQL+AW2024.032,260
BCQ2024.032,250
MOPO2024.031,900
CQL2024.03500
BC2024.0370