I received my B.Sc. in Computer Science from The Hong Kong Polytechnic University. Prior to and during my Master's studies, I worked closely with PhD candidates in the ARISE lab under the supervision of Prof. Michael R. Lyu. My recent research focus on understanding code inteligence of Multi-modal Large Language Models (MLLMs) in multimodal scenarios, and my interests revolve around Agentic AI, AI4SE & SE4AI, Social Network Analysis, and Graph Neural Networks. My existing research have also contributed to AI and strategic practices in engineering management.
I am actively seeking opportunities for Ph.D. studies in the fall of 2026!
Interests
- Multimodal LLM
- Software Engineering
- AI4SE
Education
- MSc Information EngineeringThe Chinese University of Hong Kong
- BSc Computer ScienceThe Hong Kong Polytechnic University
melvin
wang
outlook
comThe Chinese University of Hong Kong
news
selected publications

DesignBench: A Comprehensive Benchmark for MLLM-based Front-end Code Generation
J Xiao, M Wang, MH Lam, Y Wan, J Liu, Y Huo, MR Lyu
ArXiv Preprint. Link, June 2025

How Large Language Models Empower the Analysis of Online Public Engagement for Mega Infrastructure Projects: Cases in Hong Kong
M Wang, R Ma, GQ Shen, J Xue
IEEE Transactions on Engineering Management. Link, March 2025

Large Language Model-Based Data-Driven Framework for Digital Transformation in the Construction Industry
R Ma, GQ Shen, P Lou, M Wang
ASCE International Conference on Computinng in Civil Engineering, Pittsburgh, USA. Link, August 2024
experiences

Research Intern
Jan 2026 – Present | RemoteDepartment of Electrical and Computer Engineering, University of Waterloo
- Working on exploring the cost structure of MLLM educational deployment in block-based programming
- Proposing a cost-aware multimodal repair pipeline for block-based programming that uses a cheap runtime probe to estimate repair difficulty and schedule the optimal repair option
Advisor: Jialu Zhang Assistant Professor, University of Waterloo

Co-researcher
Mar 2025 – Present | Hong KongDepartment of Computer Science and Engineering, The Chinese University of Hong Kong
- Worked on analyzing code intelligence of Multi-modal LLMs in various front-end programming scenarios, contributing to the paper Designbench
- Proposed a multi-framework, multi-task evaluation benchmark for assessing MLLMs' capabilities in automated front-end engineering
- Data Collection and Processing, Metrics, Evaluation, Benchmarking
Advisor: Michael R. Lyu Professor, The Chinese University of Hong Kong

Research Assistant
Sep 2024 – Present | Hong KongDepartment of Building and Real Estate, The Hong Kong Polytechnic University
- Worked on LLM-driven Online Social Network Analysis for Mega infrastructure Project Stakeholder Analysis. Published a Journal paper in IEEE-TEM as the 1st author; co-authored a conference paper, which proposed a LLM-based data-driven framework for digital transformation in civil engineering
- Data Retrieval, Large Language Model, Social Network Analysis, Paper Completion and revision
Advisor: Jin Xue Lecturer, The University of Sydney



