Offline Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
848WW to WD
5
Feb 26, 2026
24.6Normalized Return
5
Feb 26, 2026
45.7HalfCheetah (medium-replay)
5
Feb 26, 2026
1.43Avg True Return
5
Feb 26, 2026
94.9Kitchen Complete Rate
5
Feb 26, 2026
100Success Rate
5
Feb 26, 2026
2,660.3Average Return
5
Feb 26, 2026
4,016.6Average Return
5
Feb 26, 2026
83.3Normalized Score
5
Feb 26, 2026
55.6Normalized Score
5
Feb 26, 2026
100Average Normalized Score
5
Feb 26, 2026
35.4HalfCheetah Return (Random)
5
Feb 26, 2026
2.5Gamer-Normalized Score
5
Feb 26, 2026
111.9Gamer-Normalized Score
5
Feb 26, 2026
104.2Gamer-normalized Score
5
Feb 26, 2026
267.5Gamer-normalized Score
5
Feb 26, 2026
78.7Normalized Score
4
May 28, 2026
0.91NAUC
4
May 27, 2026
0.82NAUC
4
May 27, 2026
78NAUC
4
May 27, 2026
82NAUC
4
May 27, 2026
0.75NAUC
4
May 27, 2026
56NAUC
4
May 27, 2026
95NAUC
4
May 27, 2026
93NAUC
4
May 27, 2026