not

Phish or Legit: Do LLMs Know When NOT to Cry Wolf?

Kaggle Benchmarking Challenge Submission by Anio Joseph What I Benchmarked Most security benchmarks ask: "Can the model detect the threat?" That's the easy part.