CVE-Bench

CVE-Bench is a real-world cybersecurity benchmark that contains critical-severity Common Vulnerabilities and Exposures, used to evaluate the ability of AI agents to exploit web application vulnerabilities in scenarios that mimic real-world conditions.

Papers using CVE-Bench (1)

CVE-Bench: A Benchmark for AI Agents' Ability to Exploit Real-World Web Application Vulnerabilities2025 · 1 cites