Bowen Yu (郁博文)

PhD Student in Data Science
City University of Hong Kong

📧 Email 💻 GitHub 🔗 ORCID

About Me

I am a first-year PhD student in Data Science at City University of Hong Kong, advised by Prof. Xiangyu Zhao in the Applied Machine Learning Lab. My research focuses on scalable reasoning in large language models, with emphasis on chain-of-thought compression and efficient inference for mathematical reasoning.

Before starting my PhD, I completed an MSc in Venture Creation at CityU and a BEng in Electrical Engineering and Automation at Nantong University. My research combines tensor methods, reinforcement learning, and generative AI to build models that reason reliably under computational constraints.

Research Interests

Publications

Renormalization Group Guided Tensor Network Structure Search Accepted
Maolin Wang, Bowen Yu (co-first authors), et al.
AAAI Conference on Artificial Intelligence (AAAI 2026)

Physics-inspired multi-scale framework that achieves state-of-the-art compression ratios while being 4-600× faster than existing tensor network methods.

Deep Tensor Factorization Revisited: Translation Invariance and Frequency-Limited Priors Under Review
Maolin Wang, Bowen Yu (co-first authors), et al.
IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR 2026)

Unified framework integrating dimension-specific convolutions and frequency priors for robust tensor completion across multiple modalities including hyperspectral images, fMRI, video, audio, and seismic data.

Education

News