Assistant Professor, School of Data Science, Fudan University.
My research is in machine learning and mathematical optimization. Much of my work concerns fast algorithms for large-scale graphs: locally evolving set processes, personalized PageRank, online node labeling, and graph-structured sparse optimization. I also work on the mathematical foundations of large language models, including in-context learning and the training dynamics of diffusion language models.
- NLP and LLMs, Fall 2026 (CS40008.01, Fudan)
- NLP and LLMs, Spring 2026
- Generative diffusion models
- LocalCH: iterative methods via locally evolving set processes
- FastONL: fast online node labeling
- dmo-fw: code for the ICML 2022 paper
- Graph-GHTP: Graph-IHT and Graph-GHTP
Homepage · Google Scholar · DBLP · bjzhou [at] fudan [dot] edu [dot] cn