List your company
ResearchFormerly WizWand

Reinforcement Learning on Antmaze large diverse

25OSRCal-QL
-15.7512.519.25May 11, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Cal-QLvariant=+SAC2026.05250
CQLvariant=+SAC2026.0523.30.1
RankQ2026.0521.784.7
RankQvariant=+SAC2026.0521.70
Cal-QL2026.0518.374
CQL2026.051021
Hybrid RL2026.0500
SAC2026.0500
SACvariant=+OFF2026.0500.1