List your company
ResearchFormerly WizWand

Continuous Control on Ant (Robustness Gap)

0.16Robustness GapUP(mlp(wide))
0.1460.24050.3350.4295Apr 30, 2026
Updated 6mo ago

Evaluation results

MethodLinks
UP(mlp(wide))Context=friction, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.16
UP(mlp(narrow))Context=friction, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.17
UP(lstm)Context=friction, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.21
UP(gru)Context=friction, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.21
UP(mlp(wide))Context=gravity, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.45
UP(mlp(narrow))Context=gravity, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.46
UP(gru)Context=gravity, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.49
UP(lstm)Context=gravity, State-action history length (k)=4, Data retention fraction (alpha)=Tuned2026.040.51