Share your thoughts, 1 month free Claude Pro on usSee more

Offline-to-Online Reinforcement Learning on D4RL Aggregate

77.6Average Normalized Score

Loss Smoothing

Updated 23d ago

Evaluation Results

Method	Links
Loss Smoothing 2026.07		77.6
ROAD 2026.05		71.95
OPT 2026.07		65.9
0.1 2026.05		62.15
BR 2026.05		61.8
Decreasing 2026.05		59
0.3 2026.05		58.66
Uniform 2026.05		58.45
Hard-Switch 2026.07		58.2
0.2 2026.05		57.28
0.4 2026.05		56.49
0.5 2026.05		56.26
0.0 2026.05		54.07
Cal-QL 2026.07		48.8
Online 2026.07		39.4
Offline 2026.07		27.3
AWAC 2026.07		21.7