ResearchBenchmarksNoisy Action Robust MDP on Walker2D high randomness MuJoCoFollow502.6Worst-case ReturnESPER409.6448433.7774457.91482.0426Jul 25, 2024Evaluation ResultsMethodMethodLinksWorst-case ReturnESPERnumber of seeds=5, adv...number of seeds=5, adversarial policies=8, target returns=best2024.07502.6ARDTnumber of seeds=5, adv...number of seeds=5, adversarial policies=8, target returns=best2024.07492.63DTnumber of seeds=5, adv...number of seeds=5, adversarial policies=8, target returns=best2024.07413.22