Evaluating different LLMs for their security research capabilities (opens in new tab)

A hands-on look at how frontier and open models identify, validate, reject, and exploit potential vulnerabilities differently in real-world security scans.