Share your thoughts, 1 month free Claude Pro on usSee more

MuJoCo

Benchmarks

Task Name	Dataset Name	SOTA Result
Time Series Reconstruction	MuJoCo (test)	MSE0.285	51
Offline Reinforcement Learning	MujoCo halfcheetah	Normalized Return60.8	33
Continuous Control	MuJoCo Hopper v4	Normalized Performance3,592	28
Reinforcement Learning	MuJoCo Half-Cheetah	Average Return13,907	28
Offline Reinforcement Learning	MuJoCo hopper D4RL (medium-replay)	Normalized Return101.6	26
Offline Reinforcement Learning	MuJoCo walker2d-medium D4RL	Normalized Return92.5	26
Continuous Control	MuJoCo Ant	Average Reward6,336	26
Continuous Control	MuJoCo HalfCheetah	Average Reward13,144	25
Continuous Control	MuJoCo Ant v4	Normalized Return136	24
Reinforcement Learning	MuJoCo HumanoidStandup	Average Performance130,892	24
Reinforcement Learning	MuJoCo Ant	Average Return7,889.1	24
Reinforcement Learning	MuJoCo Hopper	Average Return3,876	24
3D Dynamics Prediction	MuJoCo Fall-and-rebound scenario	Translation Error (m)0.0048	20
Offline Reinforcement Learning	MuJoCo halfcheetah-medium-replay D4RL	Normalized Return54.1	20
Offline Reinforcement Learning	MuJoCo halfcheetah-medium D4RL	Normalized Return65.6	20
Continuous Control	MuJoCo Reacher v4	Normalized Performance103	18
Continuous Control	MuJoCo Pusher v4	Normalized Performance1.36	18
Continuous Control	MuJoCo HumanoidStandup v4	Normalized Performance1.29	18
Offline Reinforcement Learning	MuJoCo halfcheetah-medium-expert D4RL	Normalized Return101.1	18
Continuous Control	MuJoCo Reacher	Average Reward6.22	18
Reinforcement Learning	MuJoCo Hopper v5	Mean Episodic Return3,268	17
HalfCheetah	Mujoco	Reward9.48	16
Ant	MuJoCo	Recovery Time (%)5.9	16
Imitation Learning	MuJoCo	Hopper Reward109.46	15
Offline Reinforcement Learning	MuJoCo walker2d-medium 10K	Score76.3	13

Showing 25 of 144 rows