Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Jailbreak detection on GPF jailbreak prompts
Loading...
97
Accuracy
FrameShield-Crit
80.36
84.68
89
93.32
Feb 23, 2026
Accuracy
F1 Score
Updated 4d ago
Evaluation Results
Method
Method
Links
Accuracy
F1 Score
FrameShield-Crit
LLM Model=Qwen2-0.5B
2026.02
97
94
FrameShield-Last
LLM Model=Qwen2-0.5B
2026.02
96
92
JBShield
LLM Model=Qwen3-4B
2026.02
87
88
FrameShield-Crit
LLM Model=Qwen3-4B
2026.02
85
71
FrameShield-Last
LLM Model=Qwen3-4B
2026.02
84
68
JBShield
LLM Model=Qwen2-0.5B
2026.02
83
85
JBShield
LLM Model=Qwen2.5-7B
2026.02
83
85
FrameShield-Crit
LLM Model=Qwen2.5-7B
2026.02
83
64
FrameShield-Last
LLM Model=Qwen2.5-7B
2026.02
81
73
Feedback
Search any
task
Search any
task