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

Adversarial Prompts

Benchmarks

Task NameDataset NameSOTA ResultTrend
Concept ErasureAdversarial Prompts
Pre-ASR88.03
8
Nudity RemovalAdversarial Prompts (test)
MMA88
7
Adversarial Prompt Detection10,000 Adversarial Prompts
Detection Rate83
5
Showing 3 of 3 rows