Researcher

Diana Cai

Diana Cai is a computer scientist who develops probabilistic machine learning methods for scientific discovery. Her work spans sparse graph models, Bayesian inference under model misspecification, and experimental design for materials and spatial genomics. She received a 2019 Google PhD Fellowship in Machine Learning and coauthored a 2025 paper on score-based variational inference. Since 2026, she is an assistant professor of computer science at Cornell University.

Timeline

  1. 2016

    Edge-exchangeable graphs

    The 2016 NeurIPS paper “Edge-exchangeable graphs and sparsity” presents a graph model whose distribution stays invariant to edge order and that can exhibit sparsity.

  2. 2019

    A Google PhD fellowship

    At Princeton, Cai receives a 2019 Google PhD Fellowship in machine learning.

  3. 2021

    Finite mixtures under misspecification

    A 2021 paper by Cai, Trevor Campbell, and Tamara Broderick shows that slight model misspecification can make a Bayesian finite-mixture model’s component-count posterior diverge as data grows.

  4. 2024

    Designing materials and spatial genomics

    Her 2024 work includes a Bayesian active-learning algorithm that chooses experiments to reduce uncertainty in a material’s convex hull. She also coauthors a spatial-genomics study that proposes structured batch experimental design to choose informative tissue slices.

  5. 2025

    Fisher meets Feynman

    The 2025 paper “Fisher meets Feynman,” coauthored by Cai, introduces a weighted product-of-experts family for black-box variational inference.

  6. By 2026

    Research fellow at Flatiron

    Before her Cornell role begins in fall 2026, Cai is a research fellow in the Center for Computational Mathematics at the Flatiron Institute.

  7. By 2026

    Degrees from three universities

    By 2026, Cai holds an A.B. in computer science and statistics from Harvard and an M.S. in statistics from the University of Chicago. She also holds M.A. and Ph.D. degrees in computer science from Princeton.

  8. Since 2026

    Assistant professor at Cornell

    Her assistant professorship in computer science at Cornell begins in fall 2026. There, she designs and analyzes probabilistic machine learning methods for scientific discovery.

Career

  1. Since 2026Cornell UniversityAssistant Professor
  2. Until 2026Simons FoundationResearch Fellow

Education

  1. Harvard UniversityA.B.
  2. University of ChicagoM.S.
  3. Princeton UniversityM.A. and Ph.D.

Sources and credits

Photo: We have no published photo of Diana Cai yet. Send us a free-licensed photo.

  1. Edge-exchangeable graphs and sparsity papers.nips.cc. Checked 21 October 2026.
  2. Announcement of the 2019 Fellowship Awardees and Highlights from the Google PhD Fellowship Summit research.google. Checked 21 October 2026.
  3. Finite mixture models do not reliably learn the number of components proceedings.mlr.press. Checked 21 October 2026.
  4. Probabilistic Prediction of Material Stability: Integrating Convex Hulls into Active Learning arxiv.org. Checked 21 October 2026.
  5. Optimizing the design of spatial genomic studies nature.com. Checked 21 October 2026.
  6. Fisher meets Feynman: score-based variational inference with a product of experts proceedings.neurips.cc. Checked 21 October 2026.
  7. Diana Cai | Department of Computer Science | Cornell Bowers cs.cornell.edu. Checked 21 October 2026.
  8. Diana Cai dianacai.com. Checked 21 October 2026.