# VGX: Large-Scale Sample Generation for Boosting Learning-Based Software Vulnerability Analyses

arXiv: https://arxiv.org/abs/2310.15436 · HTML: https://mlsanity.com/paper/2310.15436/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 24 Oct 2023.

## Built on by (most cited first, 9 of 9)

- [VulScribeR: Exploring RAG-based Vulnerability Augmentation with LLMs](https://mlsanity.com/paper/2408.04125/) (2024)
- [An Insight into Security Code Review with LLMs: Capabilities, Obstacles, and Influential Factors](https://mlsanity.com/paper/2401.16310/) (2024)
- [MLCPD: A Unified Multi-Language Code Parsing Dataset with Universal AST Schema](https://mlsanity.com/paper/2510.16357/) (2025)
- [ANVIL: Anomaly-based Vulnerability Identification without Labelled Training Data](https://mlsanity.com/paper/2408.16028/) (2024)
- [A Study on Mixup-Inspired Augmentation Methods for Software Vulnerability Detection](https://mlsanity.com/paper/2504.15632/) (2025)
- [AVIATOR: Towards AI-Agentic Vulnerability Injection Workflow for High-Fidelity, Large-Scale Code Security Dataset](https://mlsanity.com/paper/2508.20866/) (2025)
- [LLM-based Vulnerable Code Augmentation: Generate or Refactor?](https://mlsanity.com/paper/2512.08493/) (2025)
- [Willing but Unable: Separating Refusal from Capability in Code LLMs via Abliteration](https://mlsanity.com/paper/2606.05396/) (2026)
- [CyberForge: Verified Vulnerability Injection at Repository Level for Cybersecurity Agent Training](https://mlsanity.com/paper/2608.06471/) (2026)

## Abstract (arXiv)

Accompanying the successes of learning-based defensive software vulnerability analyses is the lack of large and quality sets of labeled vulnerable program samples, which impedes further advancement of those defenses. Existing automated sample generation approaches have shown potentials yet still fall short of practical expectations due to the high noise in the generated samples. This paper proposes VGX, a new technique aimed for large-scale generation of high-quality vulnerability datasets. Given a normal program, VGX identifies the code contexts in which vulnerabilities can be injected, using a customized Transformer featured with a new value-flowbased position encoding and pre-trained against new objectives particularly for learning code structure and context. Then, VGX materializes vulnerability-injection code editing in the identified contexts using patterns of such edits obtained from both historical fixes and human knowledge about real-world vulnerabilities. Compared to four state-of-the-art (SOTA) baselines (pattern-, Transformer-, GNN-, and pattern+Transformer-based), VGX achieved 99.09-890.06% higher F1 and 22.45%-328.47% higher label accuracy. For in-the-wild sample production, VGX generated 150,392 vulnerable samples, from which we randomly chose 10% to assess how much these samples help vulnerability detection, localization, and repair. Our results show SOTA techniques for these three application tasks achieved 19.15-330.80% higher F1, 12.86-19.31% higher top-10 accuracy, and 85.02-99.30% higher top-50 accuracy, respectively, by adding those samples to their original training data. These samples also helped a SOTA vulnerability detector discover 13 more real-world vulnerabilities (CVEs) in critical systems (e.g., Linux kernel) that would be missed by the original model.