Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Safety Classification on Sorry-Bench (test)
Loading...
93.9
Accepted HF Accuracy
Latent Guard
35.036
50.318
65.6
80.882
Jul 16, 2025
Accepted HF Accuracy
Updated 18d ago
Evaluation Results
Method
Method
Links
Accepted HF Accuracy
Latent Guard
Base Model=LLAMA2-CHAT-7B
2025.07
93.9
Llama Guard 3
Base Model=QWEN2-INSTR...
2025.07
59.4
Latent Guard
Base Model=LLAMA3-INST...
2025.07
59.3
Latent Guard
Base Model=QWEN2-INSTR...
2025.07
54.6
Llama Guard 3
Base Model=LLAMA2-CHAT-7B
2025.07
45.5
Llama Guard 3
Base Model=LLAMA3-INST...
2025.07
37.3
Feedback
Search any
task
Search any
task