Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Safety on Beavertails

1.2Violation Rate

Vanilla SFT

-1.908819.075640.0661.0444Feb 2, 2026
Updated 4d ago

Evaluation Results

MethodLinks
2026.02
1.2-
2026.02
1.9-
2026.02
4.8-
2026.02
17.1-
18.6-
2026.02
39.9-
2026.02
40-
2026.02
43.37-
2026.02
45.78-
2026.02
47-
2026.02
47.59-
2026.02
48.8-
2026.02
49.39-
2026.02
50.6-
2026.02
53.01-
2026.02
53.61-
2026.02
54.21-
2026.02
54.82-
2026.02
55.1-
2026.02
57.83-
2026.02
57.83-
2026.02
59.03-
2026.02
59.04-
2026.02
63.86-
2026.02
63.86-
2026.02
64.46-
2026.02
66.27-
2026.02
66.27-
2026.02
70.48-
2026.02
71.08-
2026.02
76.51-
2026.02
78.92-
2025.05
-76
2025.05
-79.57
2025.05
-96
2025.05
-94.14
2025.05
-87.95
2025.05
-97
2025.05
-96
2025.05
-97.04