I'm a second-year MEng Computer Science and Mathematics student at University College London (UCL). I'm interested in the mathematical and computational foundations of intelligent systems, especially machine learning, agentic AI, and the reliability and safety of AI systems.

My work so far spans lightweight computer vision, AI-powered applications, and research. Alongside computing and mathematics, I enjoy linguistics, literature, and the ways technology interacts with society.

• 🎓 Studying: MEng Computer Science and Mathematics, UCL
• 🔬 Interested in: Machine learning, AI systems, multi-agent systems, verification and safety
• 🛠️ Building with: Python, JavaScript/TypeScript, Vue, Node.js, and LLM tooling
• 🌏 Languages: Chinese (native), English, Japanese (learning)

Published work

Pu, H. & Yi, K. (2024). A Comparative Analysis of EfficientNet and MobileNet Models’ Performance on Limited Datasets: An Example of American Sign Language Alphabet Detection. Highlights in Science, Engineering and Technology, 94, 558–564. DOI: 10.54097/yh5d3s04

Ongoing work

• Biomedical NLP Systematic Review — Contributing as an independent second reviewer to a systematic review in Biomedical NLP.

• Institutional Design for Multi-Agent Systems (early-stage independent research) — Exploring the question: Can institutional architectures improve the robustness of multi-agent systems? The idea focuses on how communication protocols, coordination rules, oversight, and mechanisms for handling unreliable agents may affect system-level reliability and safety. Currently at the research-question and experimental-design stage.

I enjoy teaching, curriculum design, language learning, and interdisciplinary work. Through student activities and education projects, I've explored how to make complex ideas more accessible and how people learn across cultural and disciplinary boundaries.

• 🌐 Website: hongwen-pu.cn
• ✉️ Email: boypu123@outlook.com
• 💼 Feel free to reach out about AI systems, research, software projects, or collaboration.