List your company
ResearchFormerly WizWand

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

156.4Normalized ScoreA2PO
-6.77635.58777.95120.313Mar 12, 2024
Updated 8mo ago

Evaluation results

MethodLinks
A2PO2024.03156.4
TD3+BC2024.03128.5
Diffusion-QL2024.03116.3
LAPO2024.0369.7
EQL2024.0357
IQL2024.0345.7
AWAC2024.0343.9
BCQ2024.0343
CQL2024.0312.5
CQL+AW2024.0310.3
BC2024.031.1
MOPO2024.03-0.5