Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Detection on MAS LLaMA-3.1-8B (Current)
Loading...
99.15
Accuracy
AgentLens
66.7748
75.1799
83.585
91.9901
Jun 21, 2026
Accuracy
F1 Score
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
F1 Score
AgentLens
Method Category=Interp...
2026.06
99.15
99.17
Gemini-3-Pro
Method Category=LLM-as...
2026.06
96.16
96.28
Claude Sonnet-4.6
Method Category=LLM-as...
2026.06
95.31
95.58
LLaMA-Guard-4-12B
Method Category=Guardr...
2026.06
94.03
93.86
Deepseek-V4-Flash
Method Category=LLM-as...
2026.06
93.18
93.31
Claude Opus-4.6
Method Category=LLM-as...
2026.06
92.96
93.52
Deepseek-V4-Pro
Method Category=LLM-as...
2026.06
91.04
91.18
AGrail (GPT-4o)
Method Category=Guardr...
2026.06
82.09
84.55
GPT-4o
Method Category=LLM-as...
2026.06
78.04
82.02
GPT-5.2
Method Category=LLM-as...
2026.06
68.02
76.27
Feedback
Search any
task
Search any
task