Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Accidental Disclosure on CFH-Hard Accidental
Loading...
89
Accuracy (CFH-Hard Accidental)
Undefended
-3.56
20.47
44.5
68.53
Oct 20, 2025
Accuracy (CFH-Hard Accidental)
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy (CFH-Hard Accidental)
Undefended
2025.10
89
ACF
2025.10
89
LP
2025.10
89
LlamaFirewall
Backbone=o4-mini
2025.10
89
LlamaFirewall
Backbone=llama
2025.10
83
LlamaFirewall
Backbone=4o-mini
2025.10
78
LlamaFirewall
Backbone=4o
2025.10
50
CONTROLVALVE
2025.10
0
Feedback
Search any
task
Search any
task