About me

I am a Project Researcher at the University of Tokyo in the Matsuo Laboratory. Previously, I was a Ph.D. student in computer science working in Guido Montúfar’s group at Max Planck Institute for Mathematics in the Sciences (MPI MiS) in Leipzig, Germany.

I am broadly interested in machine learning, focusing more on the mathematical aspects. I am particularly fascinated by deep learning theory for its benefits for understanding deep learning mechanics, knowledgeably designing machine learning models, and ensuring positive societal impacts.

My research interests include:

  • expressivity, generalization, and optimization;
  • over-parameterized networks and implicit bias;
  • realistic settings such as few-shot learning and continual learning;
  • network architectures, such as transformers, graph neural networks, VAEs, and Bayesian neural networks.