CWE-bench v2: More coverage, quality, and transparency
Hugging Face
Hugging Face announces CWE-bench v2, offering increased coverage, quality, and transparency.
Hugging Face has released CWE-bench v2, an updated benchmark designed to provide greater coverage, improved quality, and enhanced transparency. This version aims to build upon previous iterations to better serve developers and researchers working on code security and evaluation. The update addresses prior limitations by expanding its scope and evaluation metrics.
Why it matters: The release of CWE-bench v2 provides the AI and developer community with an improved tool for evaluating code security and software vulnerabilities.
