## **PoC** [PEVuln](https://github.com/nross12/PEVuln) contains 45,255 samples of of malware. - Using the EMBER features. by [Nathan Ross](https://pure.qub.ac.uk/en/persons/nathan-ross) ## **Details** Including information on ATT&CK techniques, CWE, Vulnerability payloads and more. A vulnerable sample contains epxloitable code path or behavior with PoC. via Malvuln and ExploitDB. The dataset has a great deal of information on things like Packer Complexity, notably UPX was the most common packer used representing about half the samples. > The most common bug in malware samples is permission issues, followed by hidden functionality, hardcoded credentials, and improper authentication. [paper](https://pure.qub.ac.uk/en/publications/pevuln-a-benchmark-dataset-for-using-machine-learning-to-detect-v)