Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Prompt Injection Defense on Jailbreak APGD
Loading...
1
ASR
Gradient-guided Token Suppression
-2.84
23.08
49
74.92
May 24, 2026
ASR
Updated 8d ago
Evaluation Results
Method
Method
Links
ASR
Gradient-guided Token Suppression
Perturbation limit (ε)=16
2026.05
1
CIDER
Perturbation limit (ε)=16
2026.05
6
GUARD
Perturbation limit (ε)=16
2026.05
6
GUARD
Perturbation limit (ε)=32
2026.05
6
Gradient-guided Token Suppression
Perturbation limit (ε)=32
2026.05
6
ECSO
Perturbation limit (ε)=32
2026.05
13
ECSO
Perturbation limit (ε)=16
2026.05
16
CIDER
Perturbation limit (ε)=32
2026.05
27
ASTRA
Perturbation limit (ε)=16
2026.05
29
ASTRA
Perturbation limit (ε)=32
2026.05
32
W/O DEFENSE
Perturbation limit (ε)=16
2026.05
80
W/O DEFENSE
Perturbation limit (ε)=32
2026.05
97
Feedback
Search any
task
Search any
task