Kaiwen Shi Research

Oliver Lab, Center for AI in Protein Dynamics, Vanderbilt University.

prof_pic.jpg

I am Kaiwen Shi, a graduate of Vanderbilt University. I am interested in general mathematics and how it manifests in machine learning algorithms. I am also particularly interested in how machine learning and deep learning could benefit biomedical researches such as drug discovery. I have been working with Dr. Soheil Kolouri and Dr. Akram Aldroubi on Optimal Transport Theory. Recently, I joined OliverLab and started to work on Motif Mining problem with Dr. Carlos Oliver.

If you want to see more on my personal bio, please visit my LinkedIn page. Contact me anytime.

news

Nov 10, 2026 I passed my PhD preliminary exam at Vanderbilt CS. Onward to candidacy!
Sep 18, 2026 Two papers accepted to the NeurIPS 2026 Main Track as posters: Ensembits: an alphabet of protein conformational ensembles and BioBlobs: Unsupervised Discovery of Functional Substructures for Protein Function Prediction. See you in Atlanta!
Jul 11, 2025 PhD admission to Vanderbilt CS Department, starting this fall.
May 15, 2025 I am pleased to announce that I will join OliverLab over the summer to work on motif mining problem.
Feb 11, 2025 Our paper Linear Spherical Sliced Optimal Transport: A Fast Metric for Comparing Spherical Data get selected as an ICLR Spotlight.

selected publications

  1. Linear Spherical Sliced Optimal Transport: A Fast Metric for Comparing Spherical Data
    Xinran Liu , Yikun Bai , Rocio Diaz Martin , and 5 more authors
    In The Thirteenth International Conference on Learning Representations , 2025
  2. Ensembits: an alphabet of protein conformational ensembles
    Kaiwen Shi , and Carlos Oliver
    In Advances in Neural Information Processing Systems (NeurIPS) , 2026
  3. BioBlobs: Unsupervised Discovery of Functional Substructures for Protein Function Prediction
    Xin (Allen) Wang , Kaiwen Shi , and Carlos Oliver
    In Advances in Neural Information Processing Systems (NeurIPS) , 2026