Offline Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
93.3Success Rate (medium play)
5
May 26, 2026
96AntMaze Medium Explore Success Rate
5
May 18, 2026
56Cube Triple Play Success Rate
5
May 18, 2026
13Success Rate
5
May 12, 2026
14Success Rate
5
May 12, 2026
77.1Total Average Score
5
May 8, 2026
69.1Humanoid Score
5
May 8, 2026
78.3Humanoid Score
5
May 8, 2026
50Humanoid Score
5
May 8, 2026
87.5Success Rate (umaze)
5
Apr 6, 2026
944.46Performance (Top Left)
5
Mar 18, 2026
854.5Stand Score
5
Mar 18, 2026
257.59Run Score
5
Mar 18, 2026
902.63Stand Score
5
Mar 18, 2026
124.8U-Maze Score
5
Mar 17, 2026
9.5Average Normalized Score (All)
5
Mar 5, 2026
105.2Score
5
Feb 26, 2026
127.5Score
5
Feb 26, 2026
68.1Normalized Return
5
Feb 26, 2026
93.2Normalized Return
5
Feb 26, 2026
54.8Normalized Return
5
Feb 26, 2026
56.6Normalized Return
5
Feb 26, 2026
36.6Normalized Return
5
Feb 26, 2026
73.2Normalized Return
5
Feb 26, 2026
57.4Normalized Return
5
Feb 26, 2026