Druv Pai

druvpai AT berkeley DOT edu


Hi! I’m a PhD student in EECS at UC Berkeley, where I am fortunate to be advised by Prof. Yi Ma. Prior to this, I completed a BA in CS and MS in EECS at UC Berkeley, finishing both degrees with a 4.0 GPA.

My research interests broadly lie in simplifying deep learning. More specifically, I am interested in the following research topics:

  • geometry and statistics of latent representations produced by modern deep learning methods.
  • connecting modern deep learning practice to classical signal processing and statistics theory.
  • leveraging theoretical knowledge to design interpretable and principled learning algorithms.

I’m currently working on theories of representation learning and optimization for diffusion models and transformer networks. My recent work has been on other forms of theoretical representation learning. I have also worked in game-theoretic modeling, especially of multi-agent systems, and the intersection between learning and games.

In my free time, I play basketball and chess, and read sci-fi novels.

Notes for undergraduates and masters students.

Note 1: I am happy to chat about my research or general advising. Please send me an email and we can work out a time.

Note 2: If you are interested in research collaboration, please send me an email with your background and specific interests (the more detailed, the better). The recommended time investment is at least 15 hours per week. Unfortunately, right now my schedule is tight and generally does not permit consistent long-term mentoring of younger students, so some degree of self-sufficiency would be highly valued. To ensure a more fruitful collaboration, it would be best to have the technical knowledge to read and understand deep learning papers, especially theory-oriented work. Thank you for your understanding.

selected publications

  1. White-Box Transformers via Sparse Rate Reduction
    Yaodong YuSam BuchananDruv Pai, and 5 more authors
  2. Representation Learning via Manifold Flattening and Reconstruction
    Michael PsenkaDruv PaiVishal Raman, and 2 more authors
  3. Pursuit of a Discriminative Representation for Multiple Subspaces via Sequential Games
    Druv PaiMichael PsenkaChih-Yuan Chiu, and 3 more authors
    Journal of the Franklin Institute, 2022