Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Defense against adversarial prompts on Ring-A-Bell (test)
Loading...
0
Attack Success Rate
GLoCE
-3.9252
22.5699
49.065
75.5601
Jul 1, 2026
Attack Success Rate
Updated 23d ago
Evaluation Results
Method
Method
Links
Attack Success Rate
GLoCE
Base Model=SDv1.5, Eva...
2026.07
0
AdvUnlearn
Base Model=SDv1.5, Eva...
2026.07
0.93
DES
Base Model=SDv1.5, Eva...
2026.07
0.93
SAGE
Base Model=SDv1.5, Eva...
2026.07
1.87
GuardT2I
Base Model=SDv1.5, Eva...
2026.07
3.1
SalUn
Base Model=SDv1.5, Eva...
2026.07
3.74
OpenAI
Base Model=SDv1.5, Eva...
2026.07
19.26
UCE
Base Model=SDv1.5, Eva...
2026.07
21.5
ESD
Base Model=SDv1.5, Eva...
2026.07
26.17
Latent Guard
Base Model=SDv1.5, Eva...
2026.07
43.1
Microsoft
Base Model=SDv1.5, Eva...
2026.07
44.02
SAFREE
Base Model=SDv1.5, Eva...
2026.07
76.64
SPM
Base Model=SDv1.5, Eva...
2026.07
91.59
SLD-strong
Base Model=SDv1.5, Eva...
2026.07
97.2
SDv1.5
Base Model=SDv1.5, Eva...
2026.07
98.13
Feedback
Search any
task
Search any
task