Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

VPI

Benchmarks

Task NameDataset NameSOTA ResultTrend
Backdoor Attack DetectionVPI
Variance0
19
Text GenerationVPI Generation Tasks Llama3-8B Mistral-7B (test)
ASR100
16
Prompt Injection and Jailbreak DetectionVPI
Recall100
3
Showing 3 of 3 rows