Safety Evaluation
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Multiple-choice safety evaluation
84.48Refusal Rate
11
Safety Evaluation
73.97Overall Safety Score
9
Safety Evaluation Text-only
0.4Training Time (hr)
5
Safety Evaluation Multimodal
1Inference Overhead (Multiplier)
4
Safety Evaluation
0Illegal Content Count
3
Safety Evaluation Strict Safety Mode
50Precision
1