List your company
ResearchFormerly WizWand

Offline Reinforcement Learning on Walker2d kinematic shifts

106ScoreDROCO
40.4857.4974.591.51Dec 2, 2025
Updated 8mo ago

Evaluation results

MethodLinks
DROCOSource Policy=expert, Target Policy=expert2025.12106
OTDFSource Policy=expert, Target Policy=expert2025.1298.9
DROCOSource Policy=medium, Target Policy=expert2025.1294.6
IGDFSource Policy=expert, Target Policy=expert2025.1293.7
IQLSource Policy=expert, Target Policy=expert2025.1290.1
DROCOSource Policy=expert, Target Policy=medium2025.1283
IGDFSource Policy=medium, Target Policy=expert2025.1282.5
OTDFSource Policy=medium, Target Policy=expert2025.1276.8
IQLSource Policy=medium, Target Policy=expert2025.1271.4
DROCOSource Policy=medium, Target Policy=medium2025.1270.8
IGDFSource Policy=expert, Target Policy=medium2025.1258.6
OTDFSource Policy=expert, Target Policy=medium2025.1257.9
IQLSource Policy=expert, Target Policy=medium2025.1255.4
IGDFSource Policy=medium, Target Policy=medium2025.1251.8
IQLSource Policy=medium, Target Policy=medium2025.1248.7
OTDFSource Policy=medium, Target Policy=medium2025.1243