List your company
ResearchFormerly WizWand

Offline-to-Online Reinforcement Learning on D4RL 6 environments min-max normalized (averaged)

0.031Normalized RegretSMAC
-0.006240.245130.49650.74787Feb 19, 2026
Updated 8mo ago

Evaluation results

MethodLinks
SMACOnline Algorithm=SAC2026.020.031
SMACOnline Algorithm=TD32026.020.09
SMACOnline Algorithm=TD3+BC2026.020.226
SMACOnline Algorithm=AWR2026.020.38
CalQL/CQLOnline Algorithm=TD32026.020.442
CalQL/CQLOnline Algorithm=SAC2026.020.448
IQLOnline Algorithm=SAC2026.020.471
CalQL/CQLOnline Algorithm=AWR2026.020.482
IQLOnline Algorithm=TD3+BC2026.020.494
IQLOnline Algorithm=AWR2026.020.508
TD3+BCOnline Algorithm=TD32026.020.545
TD3+BCOnline Algorithm=TD3+BC2026.020.562
CalQL/CQLOnline Algorithm=TD3+BC2026.020.614
IQLOnline Algorithm=TD32026.020.653
TD3+BCOnline Algorithm=AWR2026.020.654
TD3+BCOnline Algorithm=SAC2026.020.962