arXiv subjects

Gary A. McCully

Publications and source records attributed to Gary A. McCully.

4 recordsLinked to original sources

arXiv2025

Comparing Unidirectional, Bidirectional, and Word2vec Models for Discovering Vulnerabilities in Compiled Lifted Code

Ransomware and other forms of malware cause significant financial and operational damage to organizations by exploiting long-standing and often difficult-to-detect software vulnerabilities. To detect vulnerabilities such as buffer overflows in compiled code, this research investigates the application of unidirectional transformer-based embeddings, specifically GPT-2. Using a dataset of LLVM functions, we trained a GPT-2 model to generate embeddings, which were subsequently used to build LSTM neural networks to differentiate between vulnerable and non-vulnerable code. Our study reveals that embeddings from the GPT-2 model significantly outperform those from bidirectional models of BERT and RoBERTa, achieving an accuracy of 92.5% and an F1-score of 89.7%. LSTM neural networks were developed with both frozen and unfrozen embedding model layers. The model with the highest performance was achieved when the embedding layers were unfrozen. Further, the research finds that, in exploring the impact of different optimizers within this domain, the SGD optimizer demonstrates superior performance over Adam. Overall, these findings reveal important insights into the potential of unidirectional transformer-based approaches in enhancing cybersecurity defenses.

cs.CR↗

arXiv2025

Impact of Data Snooping on Deep Learning Models for Locating Vulnerabilities in Lifted Code

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 unidirectional and bidirectional transformer-based embeddings. The research specifically focuses on how model performance is affected when embedding models are trained with datasets, which include samples used for neural network training and validation. The results show that introducing data snooping did not significantly alter model performance, suggesting that data snooping had a minimal impact or that samples randomly dropped as part of the methodology contained hidden features critical to achieving optimal performance. In addition, the findings reinforce the conclusions of previous research, which found that models trained with GPT-2 embeddings consistently outperformed neural networks trained with other embeddings. The fact that this holds even when data snooping is introduced into the embedding model indicates GPT-2's robustness in representing complex code features, even under less-than-ideal conditions.

cs.CR↗

arXiv2024

Bi-Directional Transformers vs. word2vec: Discovering Vulnerabilities in Lifted Compiled Code

Detecting vulnerabilities within compiled binaries is challenging due to lost high-level code structures and other factors such as architectural dependencies, compilers, and optimization options. To address these obstacles, this research explores vulnerability detection using natural language processing (NLP) embedding techniques with word2vec, BERT, and RoBERTa to learn semantics from intermediate representation (LLVM IR) code. Long short-term memory (LSTM) neural networks were trained on embeddings from encoders created using approximately 48k LLVM functions from the Juliet dataset. This study is pioneering in its comparison of word2vec models with multiple bidirectional transformers (BERT, RoBERTa) embeddings built using LLVM code to train neural networks to detect vulnerabilities in compiled binaries. Word2vec Skip-Gram models achieved 92% validation accuracy in detecting vulnerabilities, outperforming word2vec Continuous Bag of Words (CBOW), BERT, and RoBERTa. This suggests that complex contextual embeddings may not provide advantages over simpler word2vec models for this task when a limited number (e.g. 48K) of data samples are used to train the bidirectional transformer-based models. The comparative results provide novel insights into selecting optimal embeddings for learning compiler-independent semantic code representations to advance machine learning detection of vulnerabilities in compiled binaries.

cs.CR↗

arXiv2024

Confronting the Reproducibility Crisis: A Case Study of Challenges in Cybersecurity AI

In the rapidly evolving field of cybersecurity, ensuring the reproducibility of AI-driven research is critical to maintaining the reliability and integrity of security systems. This paper addresses the reproducibility crisis within the domain of adversarial robustness -- a key area in AI-based cybersecurity that focuses on defending deep neural networks against malicious perturbations. Through a detailed case study, we attempt to validate results from prior work on certified robustness using the VeriGauge toolkit, revealing significant challenges due to software and hardware incompatibilities, version conflicts, and obsolescence. Our findings underscore the urgent need for standardized methodologies, containerization, and comprehensive documentation to ensure the reproducibility of AI models deployed in critical cybersecurity applications. By tackling these reproducibility challenges, we aim to contribute to the broader discourse on securing AI systems against advanced persistent threats, enhancing network and IoT security, and protecting critical infrastructure. This work advocates for a concerted effort within the research community to prioritize reproducibility, thereby strengthening the foundation upon which future cybersecurity advancements are built.

cs.LG↗