Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Detection on MAS Gemma-2-9B (Current)
Loading...
95.35
Accuracy
AgentLens
60.2812
69.3856
78.49
87.5944
Jun 21, 2026
Accuracy
F1 Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
F1 Score
AgentLens
Method Category=Interp...
2026.06
95.35
86.67
Gemini-3-Pro
Method Category=LLM-as...
2026.06
89.53
75.68
LLaMA-Guard-4-12B
Method Category=Guardr...
2026.06
88.37
58.33
Claude Sonnet-4.6
Method Category=LLM-as...
2026.06
87.21
71.79
Deepseek-V4-Flash
Method Category=LLM-as...
2026.06
87.21
71.79
Deepseek-V4-Pro
Method Category=LLM-as...
2026.06
87.21
70.27
Claude Opus-4.6
Method Category=LLM-as...
2026.06
86.05
70
AGrail (GPT-4o)
Method Category=Guardr...
2026.06
80.23
62.22
GPT-4o
Method Category=LLM-as...
2026.06
63.95
47.46
GPT-5.2
Method Category=LLM-as...
2026.06
61.63
45.9
Feedback
Search any
task
Search any
task