Hi, I’m Arastoo Zibaeirad.
I’m a Ph.D. student working at the intersection of LLMs, software, and security. My main research focuses on using LLMs for software vulnerability detection, with broader interests in adversarial machine learning and automated program repair.
About Me
I am a Ph.D. student in Computer Science at UNC Charlotte, advised by Prof. Marco Vieira. My research focuses on using Large Language Models (LLMs) for software security — with an emphasis on vulnerability detection, automated program repair, and dataset development for these tasks.
I am currently building high-quality, task-specific datasets that enable more effective training and evaluation of LLMs for software vulnerability detection and patch generation. My broader background spans network security, software security, and adversarial machine learning, and I apply LLMs to core software-engineering and cybersecurity problems such as static code analysis, patch synthesis, and secure code generation.
I love connecting with other tech enthusiasts and professionals. If you’re interested in collaboration, have a challenging question related to my research, or are working on similar security problems, feel free to reach out!
Research Interests
News
-
Dec 2025
New paper accepted — Diverse LLMs vs. Vulnerabilities: Who Detects and Fixes Them Better? at the 3rd LLM4Code Workshop, co-located with ICSE 2026.
-
Mar 2025
New preprint out — Reasoning with LLMs for Zero-Shot Vulnerability Detection .
-
Sep 2024
Released VulnLLMEval , a benchmark for evaluating LLMs on software vulnerability detection & patching.
-
Jul 2024
Published a comprehensive survey on Smart Grid security .
-
2024
Started Ph.D. at UNC Charlotte, advised by Prof. Marco Vieira.
Selected Publications
2 papers
Diverse LLMs vs. Vulnerabilities: Who Detects and Fixes Them Better?
Accepted at the 3rd International Workshop on Large Language Models for Code (LLM4Code), co-located with ICSE 2026
Reasoning with LLMs for Zero-Shot Vulnerability Detection
arXiv preprint arXiv:2503.17885 · 2025
| Apr 28, 2025 | How LLMs Reason About Code Vulnerabilities |
|---|