Previously, I was a PhD student advised by Prof. Russell Impagliazzo. My primary research interests included learning theory and complexity theory, including replicable learning algorithms and boosting.

My dissertation, "Replicable Learning Algorithms", is available online for free: PDF.
The introduction (PDF pages 20-25) is written for general audiences. It explains the dissertation's contents, and reading it does not require a mathematical background.

In October 2024, I joined the Max Planck Institute for Informatics, as a postdoc in the Algorithms and Complexity Department!

Publications

DBLP   Arxiv

• Dissertation'24 "Replicable Learning Algorithms", R. Lei. UC San Diego. ProQuest ID:Lei_ucsd_0033D_23851. Merritt ID:ark:/13030/m50d65kg. 2024. PDF.
• STOC'23 "Stability is Stable: Connections between Replicability, Privacy, and Adaptive Generalization". M. Bun, M. Gaboardi, M. Hopkins, R. Impagliazzo, R. Lei, T. Pitassi, S. Sivakumar, J. Sorrell. Proceedings of the 55th Annual ACM SIGACT Symposium on Theory of Computing
  Arxiv Version (Full Paper)
  Also accepted in the non-archival track of FORC 2023
  Selected for an oral presentation at TPDP 2023
• STOC'22 "Reproducibility in Learning". R. Impagliazzo, R. Lei, T. Pitassi, J. Sorrell. Proceedings of the 54th Annual ACM SIGACT Symposium on Theory of Computing
  Our Arxiv Version is slightly more up-to-date; namely, it includes the "reproducibility" to "replicability" name change.
  An earlier version of this work appeared in TPDP 2021
• COLT'21 "Boosting in the Presence of Massart Noise". I. Diakonikolas, R. Impagliazzo, D. M. Kane, R. Lei, J. Sorrell, C. Tzamos. Proceedings of the 34th Annual Conference on Learning Theory
  Arxiv Version
  

Note: In the theoretical computer science community, author order is typically alphabetical and denotes equal contribution.