Hi, I'm Shuang, a second-year PhD candidate in Cybersecurity at LIACS, Leiden University, Netherlands. My research focuses on the security of natural-language prompts in AI-assisted code generation. You can find more about our group here: LOCS.
In my first year of PhD, I focused on Cryptography and conducted a comprehensive review of preprocessing models in Multi-Party Computation. The manuscript is available here.
I am actively involved in the academic activities. I am a member of the LIACS Diversity Committee and ACCSS. I enjoy organizing events and have co-organized Alice and Eve 2024 and Ethics in Cybersecurity and AI 2025.
I am passionate about programming and proficient in C/C++, Python, and popular front-end technical stacks. Before starting my PhD, I worked as a research engineer for two years in BJCA, the national cryptography company in Beijing. I also worked as a developer for five years before that.
If you have any collaboration ideas or questions, please feel free to contact me.
Paper: Vibe coding: programming through conversation with artificial intelligence. 2025 1 What do they do, and what’s the result? 1.1 What do they do? 1.1.1 1.1.2 1.1.3 1.2 What’s the result? 1.2.1 1.2.2 1.2.3 2: Can I use any of it in my work? If yes, how? 2.1 … 2.1.1 2.1.2 2.1.3 … Vibe coding: programming through conversation with artificial intelligence
Paper: Code Gradients: Towards Automated Traceability of LLM-Generated Code. 2024 1 What do they do, and what’s the result? 1.1 What do they do? 1.1.1 1.1.2 1.1.3 1.2 What’s the result? 1.2.1 1.2.2 1.2.3 2: Can I use any of it in my work? If yes, how? 2.1 … 2.1.1 2.1.2 2.1.3 … Code Gradients: Towards Automated Traceability of LLM-Generated Code
Paper: Can LLMs Generate Higher Quality Code Than Human? An Empirical Study. 2025 1 What do they do, and what’s the result? 1.1 What do they do? 1.1.1 This paper conducts an empirical analysis of AI-generated code to assess whether LLMs can produce correct and higher-quality code than humans. It evaluates the code quality of … Can LLMs Generate Higher Quality Code Than Human
Paper: Language Models in Software Development Tasks: An Emperimental Analysis of Energy and Accuracy. 2025 1 What do they do, and what’s the result? 1.1 What do they do? They explore the trade-off between model accuracy and energy consumption, so that help developers to select a language model. They investigate the performance of 18 LLMs … Language Models in Software Development Tasks: An Emperimental Analysis of Energy and Accuracy
Paper: Prompt Engineering or Fine-Tuning: An Empirical Assessment of LLMs for Code. 2025 1 What do they do, and what’s the result? 1.1 What do they do? 1. Two primary methodologies are used in LLMs for code: prompt engineering and fine-tuning. In this paper, they conduct a comparative analysis between these two methodologies. 2. … Prompt Engineering or Fine-Tuning: An Empirical Assessment of LLMs for Code
This blog is combing several code metrics tools which introduced in the paper ‘Can LLMs Generate Higher Quality Code Than Humans? An Empirical Study‘. Some context are referenced directly from the paper. For more details, please see the paper. 1 Code Quality Metric 1.1 Code quality Metric 1.2 Human-centric metrics 1.3 Critical code quality metrics … Code Metrics Tools or Static Analysis Tools