ResearchBenchmarksReinforcement Learning on Ant v5 (Coefficient of Variation)Reinforcement Learning on Ant v5 (Coefficient of Variation)Follow4.2Coefficient of VariationSAC0.66824.50948.3572.191Mar 12, 2026Evaluation resultsMethodMethodLinksCoefficient of VariationSACVariant=H-EARSVariant=H-EARS2026.034.2SACVariant=VanillaVariant=Vanilla2026.035.8TD3Variant=H-EARSVariant=H-EARS2026.039.6TD3Variant=VanillaVariant=Vanilla2026.0311PPOVariant=H-EARSVariant=H-EARS2026.0325.3DDPGVariant=VanillaVariant=Vanilla2026.0337.9PPOVariant=VanillaVariant=Vanilla2026.0374.7DDPGVariant=H-EARSVariant=H-EARS2026.0392.5