Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Harmful prompt detection on ViSU
Loading...
98
Precision
Detoxify (Orig)
-3.92
22.54
49
75.46
Apr 7, 2026
Precision
Recall
F1 Score
Updated 10d ago
Evaluation Results
Method
Method
Links
Precision
Recall
F1 Score
Detoxify (Orig)
2026.04
98
26
40
HyPE
2026.04
98
98
98
HyPE
Attack Type=Style Atta...
2026.04
97
67
80
Latent Guard
2026.04
79
52
63
NSFW-Classifier
2026.04
70
80
75
NSFW-Classifier
Attack Type=Style Atta...
2026.04
65
65
65
GuardT2I
2026.04
48
77
59
DiffGuard
2026.04
27
36
31
DiffGuard
Attack Type=Style Atta...
2026.04
0
0
0
Detoxify (Orig)
Attack Type=Style Atta...
2026.04
0
0
0
Latent Guard
Attack Type=Style Atta...
2026.04
0
0
0
Feedback
Search any
task
Search any
task