ResearchBenchmarksContinuous Control on Humanoid 3M v5Follow5,180.1Average Episodic RewardPDA771.9561,916.3783,060.84,205.222Mar 10, 2026Evaluation ResultsMethodMethodLinksAverage Episodic RewardPDAEnvironment steps=3M,...Environment steps=3M, Seeds=10, Tests per epoch=102026.035,180.1PPOEnvironment steps=3M,...Environment steps=3M, Seeds=10, Tests per epoch=102026.03941.5