Humanoid
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Humanoid
90,921,063Zero-Shot Reward
32
Humanoid v3
11,888Avg Final Return
26
Humanoid Randomized Task (OOD Sweep)
-3.58Reward
24
Humanoid d = 6392
1Rank
21
Humanoid 17-Dof
13,860Final Return
21
Humanoid
5,299Cumulative Reward
16
Humanoid MuJoCo v2 (evaluation)
5,078.3Action Performance (p_act=0.1)
14
Humanoid v4
5,715Reward
13
Humanoid v5
5,906.7Average Return
13
Humanoid (delta=[0.8^6, 0.5^6, 0.2^5], kappa=4.0) v5 (test)
5,620Return
12
Humanoid MuJoCo (test)
4.02Normalized Worst-Case Reward
12
Humanoid v1 (test)
93,123.84Total Score
12
Humanoid v5
5,906.7Performance Score
11
Humanoid v4
7,365.7Mean Episode Return
10
Humanoid
18.52Relative Return Improvement
10
Humanoid
669.52Objective Value
8
Humanoid v4 (test)
6,907.99Reward
8
Humanoid v5
6.3Coefficient of Variation (%)
8
Humanoid v5
5,228Average Returns
8
Humanoid
1,734.1Episodic Reward
8
Humanoid gravity v2
6,360Average Return
8
Humanoid Standup
17.6Computational Time (s)
8
Humanoid v4
4,978.5Average Cumulative Reward
7
Humanoid v4
72.3Local Optima Escape Rate
7
Humanoid
680.1Humanoid Return (p_act=0.1)
7