Yufei (Celina) Chen

Hi! I’m Yufei (Celina) Chen, an undergraduate student at the University of Toronto pursuing a Computer Science Specialist degree and an undergraduate researcher at the Vector Institute.

I am broadly interested in machine learning, with particular focus on:

• Differential privacy
• Machine unlearning
• Data preprocessing
• Representation learning

Current work

I am currently an undergraduate researcher advised by Professor Nicolas Papernot. My research examines how models can be adapted under data scarcity and sensitivity, particularly when retraining is constrained by privacy or uneven data availability. One line of my work studies differentially private data mixing, where mixtures of public datasets are learned to compensate for limited private data without consuming additional privacy budget. In parallel, I explore machine unlearning under data scarcity, investigating whether principled data reweighting can restore long-tailed structure and improve stability during unlearning.

Alongside this work, I am interested in representation learning and interpretability, with a focus on how internal representations shape model behavior after training. I am currently investigating model merging through representation alignment, studying when and how aligning internal feature spaces improves mergeability and preserves task-specific structure. A brief summary of my current progress can be found here.

Education

• University of Toronto
  B.Sc. (Undergraduate), Computer Science Specialist
  Expected graduation: 2026

This website documents my research, projects, and academic activities.