Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Response Harmfulness Detection on BeaverTails

89.9F1 Score

BeaverDam 7B

12.73232.76652.872.834Feb 3, 2026Feb 22, 2026Mar 13, 2026Apr 1, 2026Apr 20, 2026May 9, 2026May 29, 2026
Updated 2d ago

Evaluation Results

MethodLinks
2026.02
89.9-
2026.05
89.9-
2026.05
88.3-
2026.02
87.6-
2026.05
87.6-
2026.05
87.6-
2026.05
87.44-
2026.05
86.72-
2026.05
86.72-
2026.02
86.7-
2026.05
86.7-
2026.05
86.7-
2026.05
86.55-
2026.02
86.48-
2026.05
86.29-
2026.02
86.04-
2026.02
85.89-
2026.05
85.84-
2026.02
85.35-
2026.02
85.19-
2026.02
84.99-
2026.05
84.66-
2026.02
84.4-
2026.05
84.4-
2026.05
83.41-
2026.05
83.27-
2026.05
83.2-
2026.05
82.26-
2026.05
80.89-
2026.05
80.44-
2026.05
79.96-
2026.02
79.39-
2026.05
78.7-
2026.05
78.63-
2026.05
77.89-
2026.05
77.26-
2026.02
77.1-
2026.05
77.1-
2026.02
75.2-
2026.05
75.2-
2026.02
74.7-
2026.05
74.7-
2026.05
74.7-
2026.02
73.8-
2026.05
73.8-
2026.05
73.8-
2026.05
71.8-
2026.05
71.8-
2026.02
69.51-
2026.02
67.84-
2026.05
67.84-
2026.05
67.84-
2026.05
67.1-
2026.02
63.61-
2026.05
63.61-
2026.05
63.61-
2026.05
30.97-
2026.05
30.97-
2026.05
15.7-
2026.04
-83.5
2026.04
-77.1
2026.04
-83.7
2026.04
-70
2026.04
-84.3
2026.04
-80.1
2026.04
-83.8
2026.04
-68.8