Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
CarRun on Bullet-Safety-Gym (Normalized Metrics)
Loading...
99
Normalized Reward
Reference policy (πref)
-22.68
8.91
40.5
72.09
Apr 29, 2026
May 2, 2026
May 6, 2026
May 9, 2026
May 13, 2026
May 16, 2026
May 20, 2026
Normalized Reward
Normalized Cost
Updated 13d ago
Evaluation Results
Method
Method
Links
Normalized Reward
Normalized Cost
Reference policy (πref)
2026.05
99
1.89
PREFINE
2026.05
93
0.38
BC
2026.05
86
0.11
SafeDICE
2026.05
81
0.28
FISOR
2026.04
73
14
CDT+SAS
Alignment=SAS, Base Mo...
2026.04
72
39
PPL
2026.05
71
0.39
CPQ
2026.05
-18
0.06
Feedback
Search any
task
Search any
task