Safe Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
0Cost
48
Jun 9, 2026
0Cost
48
Jun 9, 2026
0Cost
48
Jun 9, 2026
2.3Cost
48
Jun 9, 2026
100Verified Success Rate (50th Percentile)
14
Feb 26, 2026
23.79Return
12
Jun 15, 2026
10.15Return
12
Jun 15, 2026
27.7Return
12
Jun 15, 2026
1.13Return
12
Jun 15, 2026
91.4Return
12
Jun 15, 2026
1,691.92Return
12
Jun 15, 2026
2,870.7Return
12
Jun 15, 2026
6,496.85Return
12
Jun 15, 2026
3,231.6Return
12
Jun 15, 2026
63.9JR Score
12
Jun 11, 2026
0Violation Rate
12
Feb 26, 2026
0Violation Rate
12
Feb 26, 2026
26.06Violation Rate
12
Feb 26, 2026
34.03Violation Rate
12
Feb 26, 2026
0.236Violation Rate
12
Feb 26, 2026
72.49Violation Rate
12
Feb 26, 2026
34.75Violation Rate
12
Feb 26, 2026
29.12Violation Rate
12
Feb 26, 2026
17.71Violation Rate
12
Feb 26, 2026
50.16Violation Rate
12
Feb 26, 2026