List your company
ResearchFormerly WizWand

Offline-to-Online Reinforcement Learning on D4RL Aggregate

77.6Average Normalized ScoreLoss Smoothing
19.46434.55749.6564.743May 14, 2026May 22, 2026May 30, 2026Jun 7, 2026Jun 15, 2026Jun 23, 2026Jul 1, 2026
Updated 4mo ago

Evaluation results

MethodLinks
Loss Smoothing2026.0777.6
ROADMixing Ratio Strategy=ROAD, Algorithm=IQL2026.0571.95
OPT2026.0765.9
0.1Mixing Ratio Strategy=Fixed (0.1), Algorithm=IQL2026.0562.15
BRMixing Ratio Strategy=BR, Algorithm=IQL2026.0561.8
DecreasingMixing Ratio Strategy=Decreasing, Algorithm=IQL2026.0559
0.3Mixing Ratio Strategy=Fixed (0.3), Algorithm=IQL2026.0558.66
UniformMixing Ratio Strategy=Uniform, Algorithm=IQL2026.0558.45
Hard-Switch2026.0758.2
0.2Mixing Ratio Strategy=Fixed (0.2), Algorithm=IQL2026.0557.28
0.4Mixing Ratio Strategy=Fixed (0.4), Algorithm=IQL2026.0556.49
0.5Mixing Ratio Strategy=Fixed (0.5), Algorithm=IQL2026.0556.26
0.0Mixing Ratio Strategy=Fixed (0.0), Algorithm=IQL2026.0554.07
Cal-QL2026.0748.8
Online2026.0739.4
Offline2026.0727.3
AWAC2026.0721.7