Lu Yin
Professor, Shenzhen University of Advanced Technology (深圳理工大学), School of Computer Science and Artificial Intelligence · Lead of the LUMI Lab
- Visiting scholar, University of Surrey
- Visiting scholar, Eindhoven University of Technology (TU/e)
- Visiting scholar, MPI‑ELLIS, Tübingen
Also reachable at l.yin@surrey.ac.uk · l.yin@tue.nl · lu.yin@tue.ellis.eu
About
Curriculum vitae →Greetings! I’m Lu. I lead the Lightweight & Universal Machine Intelligence (LUMI) lab. Before joining SUAT I was an Assistant Professor at the University of Surrey, a Postdoctoral Fellow at TU/e, and a research scientist intern at Google’s New York City office.
Feel free to reach out if you’d like to discuss anything with me :)
Research
All publications →Efficient and Scalable Foundation Models
Making large models cheaper to train, adapt and serve — sparse training, post-training pruning and layer-wise sparsity allocation, and memory-efficient optimisation that keeps quality intact at high compression ratios.
Understanding and Enhancing LLMs
Studying why large language models behave the way they do — depth and layer utilisation, normalisation and scaling choices, weight-decay and heavy-tailed structure — and turning those findings into concrete training recipes.
LLM Agents and Self-Evolution
Agents built on large language models, and the mechanisms that let them improve themselves.
Foundation Models for Vertical Domains
Bringing foundation models into practical domains — healthcare and medical imaging, urban computing, and other applied areas.
Join us
The LUMI lab is recruiting at every level:
- Research Assistants (RA)
- PhD students — jointly supervised with Beihang University (北京航空航天大学) and Nanjing University (南京大学)
- Research Assistant Professors (RAP)
- Postdocs
If our interests overlap, email me your CV and a short note on what you would like to work on.