List your company
ResearchFormerly WizWand

Offline Reinforcement Learning on HalfCheetah kinematic shifts

79.6ScoreOTDF
40.70450.80260.970.998Dec 2, 2025
Updated 8mo ago

Evaluation results

MethodLinks
OTDFSource Policy=expert, Target Policy=expert2025.1279.6
DROCOSource Policy=expert, Target Policy=expert2025.1267.4
DROCOSource Policy=expert, Target Policy=medium2025.1258.5
OTDFSource Policy=medium, Target Policy=expert2025.1258.3
DROCOSource Policy=medium, Target Policy=expert2025.1252.6
OTDFSource Policy=expert, Target Policy=medium2025.1251.7
IQLSource Policy=expert, Target Policy=expert2025.1249.7
IGDFSource Policy=expert, Target Policy=expert2025.1247.6
IQLSource Policy=medium, Target Policy=expert2025.1247.5
IQLSource Policy=expert, Target Policy=medium2025.1247.1
IGDFSource Policy=expert, Target Policy=medium2025.1246.8
IGDFSource Policy=medium, Target Policy=expert2025.1245.4
DROCOSource Policy=medium, Target Policy=medium2025.1245.3
IQLSource Policy=medium, Target Policy=medium2025.1245.2
IGDFSource Policy=medium, Target Policy=medium2025.1245.2
OTDFSource Policy=medium, Target Policy=medium2025.1242.2