Shuo Han

I am a PhD student in Computer Science at University College London, supervised by Prof. Federica Sarro and Prof. Mark Harman.

Earlier, I earned my master’s degree in Statistics and Data Science from Northwestern University, and my bachelor’s degree in Computer Science and Statistics from Boston University.

Email  /  CV  /  Scholar

Research

My research focuses on AI Software Engineering, especially the trustworthiness of large language model (LLM)-based systems. I am particularly engaged in:

• Improving the reliability and security of LLM-based systems by studying their failure modes and robustness.
• Exploring LLM-assisted development workflows, including vibe coding and AI-driven code review, to support human–AI collaboration in software engineering.

A large-scale empirical study revealing instability of LLMs under paraphrased prompts in vulnerability detection, highlighting the need for robust prompting and model refinement.

A bi-level optimization framework that integrates crowdsourced annotations with soft-labeling and optimizes them using GroupDRO to enhance robustness against out-of-distribution risks.

Academic Services

Theme powered by Jon Barron