ResearchBenchmarksBounding Gait Control on Unitree Go2 Simulated (test)Follow14.194Average Episode Reward (ρ=0)PPO-PGDLC9.8020810.9422912.082513.22271Apr 22, 2024Evaluation ResultsMethodMethodLinksAverage Episode Reward (ρ=0)Average Episode Reward (ρ=1)Average Episode Reward (ρ=2)Average Episode Reward (ρ=3)PPO-PGDLCλ=10−3, ϵ=0.001λ=10−3, ϵ=0.0012024.0414.19413.6111.5633.078PPO-PGDLCλ=10−2, ϵ=0.001λ=10−2, ϵ=0.0012024.0413.46712.37310.8279.093PPO-PGDLCλ=5 × 10−5, ϵ=0.001λ=5 × 10−5, ϵ=0.0012024.0413.45213.08511.3086.771PPO2024.0411.28611.17210.5169.041PPO-PGDLCλ=10−4, ϵ=0.001λ=10−4, ϵ=0.0012024.049.9718.5577.3996.413