ResearchBenchmarksRobotic Control on PointMassNavigateFollow20Success RatioPPO1919.52020.5Jun 19, 2026Evaluation ResultsMethodMethodLinksSuccess RatioGradient StepsAPG SpeedupEnvironment StepsPPO2026.06201,104——APG2026.06201209.2—PPO2026.0620——263,360APG2026.0620—1.37192,000