Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety on StrongReject
Loading...
99.7
Safety Rate
Vanilla SFT
35.532
52.191
68.85
85.509
Feb 2, 2026
Safety Rate
Updated 1mo ago
Evaluation Results
Method
Method
Links
Safety Rate
Vanilla SFT
Alignment Dataset=Dire...
2026.02
99.7
Vanilla SFT
Alignment Dataset=STAR...
2026.02
99.7
DGR SFT
Alignment Dataset=Dire...
2026.02
98.7
Vanilla SFT
Alignment Dataset=R1-A...
2026.02
98.1
DGR SFT
Alignment Dataset=STAR...
2026.02
98.1
DGR SFT
Alignment Dataset=R1-A...
2026.02
98.1
Qwen2.5-7B-Instruct
2026.02
94.6
s1.1-7B
2026.02
38
Feedback
Search any
task
Search any
task