Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety guardrailing on SafeEditBench
Loading...
61.7
Macro F1
PolicyShiftGuard-7B
26.132
35.366
44.6
53.834
Jul 7, 2026
Macro F1
Overall Score
Updated 18d ago
Evaluation Results
Method
Method
Links
Macro F1
Overall Score
PolicyShiftGuard-7B
Parameter Scale=7B
2026.07
61.7
69.9
PolicyShiftGuard-3B
Parameter Scale=3B
2026.07
55.1
62.4
Qwen2.5-VL-7B
Parameter Scale=7B
2026.07
48.7
32.9
SafeGuard-VL-RL-7B
Parameter Scale=7B
2026.07
45.6
52.5
GuardReasoner-VL-3B
Parameter Scale=3B
2026.07
41
51.8
GuardReasoner-VL-7B
Parameter Scale=7B
2026.07
38.3
48.2
QwenGuard-7B
Parameter Scale=7B
2026.07
32.8
38.7
LlamaGuard4-12B
Parameter Scale=12B
2026.07
30.1
19.1
ShieldGemma2-4B
Parameter Scale=4B
2026.07
27.5
36.8
Feedback
Search any
task
Search any
task