4 papers
Comparing Unidirectional, Bidirectional, and Word2vec Models for Discovering Vulnerabilities in Compiled Lifted Code
Gary A. McCully, John D. Hastings, Shengjie Xu +1
Ransomware and other forms of malware cause significant financial and operational damage to organizations by exploiting long-standing and often difficult-to-detect software vulnera…
Impact of Data Snooping on Deep Learning Models for Locating Vulnerabilities in Lifted Code
Gary A. McCully, John D. Hastings, Shengjie Xu
This study examines the impact of data snooping on neural networks used to detect vulnerabilities in lifted code, and builds on previous research that used word2vec and unidirectio…
Watertox: The Art of Simplicity in Universal Attacks A Cross-Model Framework for Robust Adversarial Generation
Zhenghao Gao, Shengjie Xu, Meixi Chen +1
Contemporary adversarial attack methods face significant limitations in cross-model transferability and practical applicability. We present Watertox, an elegant adversarial attack…
Bi-Directional Transformers vs. word2vec: Discovering Vulnerabilities in Lifted Compiled Code
Gary A. McCully, John D. Hastings, Shengjie Xu +1
Detecting vulnerabilities within compiled binaries is challenging due to lost high-level code structures and other factors such as architectural dependencies, compilers, and optimi…