Safety Prompts
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
Safety-Prompts
92Mean Accuracy
12
901 Safety Prompts (test)
4.134Average Rank
11
Safety Prompts (randomly selected 200 samples per field)
1.5Insensitivity Score
9
HRL/LRL Safety Prompts English Multi-Image v1
2ASR
6
HRL/LRL Safety Prompts English Text v1
1ASR
6