Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Jailbreak Safety Evaluation on English dataset Single-Image
Loading...
10
StrongREJECT (Perturbed)
Gemini 1.5 Pro
-0.4
2.3
5
7.7
Apr 13, 2025
StrongREJECT (Perturbed)
StrongREJECT (Unperturbed)
StrongREJECT (Cipher)
Updated 4d ago
Evaluation Results
Method
Method
Links
StrongREJECT (Perturbed)
StrongREJECT (Unperturbed)
StrongREJECT (Cipher)
Gemini 1.5 Pro
2025.04
10
13
0
Gemini 1.5 Flash
2025.04
8
39
19
Claude 3 Haiku
2025.04
5
4
2
GPT-4o
2025.04
1
10
9
Claude 3.5 Sonnet
2025.04
0
0
0
GPT-4o Mini
2025.04
0
1
10
Feedback
Search any
task
Search any
task