Safe Reinforcement Learning
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
4.188Time To Safety (TTS)
6
Jun 15, 2026
4.591Time To Safety (TTS)
6
Jun 15, 2026
4.525Time-to-Safety (TTS)
6
Jun 15, 2026
4.204Time To Safety (TTS)
6
Jun 15, 2026
3.703Time-to-Safety (TTS)
6
Jun 15, 2026
3.098Time To Safety (TTS)
6
Jun 15, 2026
1.85Normalized Reward
6
May 22, 2026
2.39Normalized Reward
6
May 22, 2026
1.23Normalized Reward
6
May 22, 2026
1.27Normalized Reward
6
May 22, 2026
1Normalized Reward
6
May 22, 2026
1.04Normalized Reward
6
May 22, 2026
34.64Return
5
Apr 7, 2026
15.58Mean Reward
5
Apr 7, 2026
0.6Mean Reward
5
Apr 7, 2026
0Σδ
5
Apr 2, 2026
21Collisions
5
Mar 26, 2026
12Collision Count
5
Mar 26, 2026
1,675Unsafe Events
5
Mar 26, 2026
739Unsafe Event Count
5
Mar 26, 2026
1,544Unsafe Events
5
Mar 26, 2026
1Constraint Violation Rate 1
5
Feb 26, 2026
459Reward
4
Jul 1, 2026
579.5Reward
4
Jul 1, 2026
25Categorical Violation Rate (%)
4
Jun 11, 2026