PeopleofSuperintelligence

Researcher

Jeffrey Negrea

Jeffrey Negrea is a statistician who works on the theory of machine learning, with research on statistical learning theory, online learning and computational statistics. In 2019 and 2020 he coauthored papers on generalization bounds, and in 2021 a NeurIPS paper on minimax optimal quantile and semi-adversarial regret. He visited the Institute for Advanced Study in 2020, was named a Rising Star in Data Science by the University of Chicago's Data Science Institute in 2021, and completed a PhD at the University of Toronto under Daniel Roy and Jeffrey Rosenthal in 2022. After a postdoc at that institute from 2022 to 2023, he joined the University of Waterloo in July 2023, and in 2025 he coauthored a paper with Alexander Terenin on Thompson sampling for online learning. He is an assistant professor at Waterloo in the Department of Statistics and Actuarial Science and a faculty affiliate at the Vector Institute.

Timeline

  1. 2019–2020

    Generalization bounds for learned predictors

    In 2019, he coauthors a paper with Mahdi Haghifam, Gintare Karolina Dziugaite, Ashish Khisti and Daniel M. Roy that tightens mutual information bounds for Stochastic Gradient Langevin Dynamics, and another with Dziugaite and Roy on generalization via derandomization for interpolating predictors. A 2020 paper with Haghifam, Khisti, Roy and Dziugaite sharpens generalization bounds through conditional mutual information and applies them to Langevin dynamics. He is also a visiting student at the Institute for Advanced Study in 2020.

  2. 2021

    Quantile and semi-adversarial regret

    In 2021, he and Blair Bilodeau, as equal-contribution authors with Nicolò Campolongo, Francesco Orabona and Daniel M. Roy, publish a NeurIPS paper on minimax optimal regret for quantile and semi-adversarial online learning. The paper uses FTRL with root-logarithmic regularizers, which can be read as variants of NormalHedge.

  3. 2021

    A Rising Star in Data Science

    In 2021, the University of Chicago's Data Science Institute recognizes him as a Rising Star in Data Science. The cohort listing announces his upcoming post as an assistant professor at the University of Waterloo and faculty affiliate at the Vector Institute.

  4. 2022

    PhD at the University of Toronto

    In 2022, he completes a PhD in Statistical Sciences at the University of Toronto as a Vanier Scholar, following an MSc there and a BMath at the University of Waterloo. His dissertation, advised by Daniel Roy and Jeffrey Rosenthal, covers approximations and scaling limits of Markov chains with applications to MCMC and approximate inference. In 2022, he is also a visiting graduate student in the Learning and Games program at the Simons Institute for the Theory of Computing.

  5. 2022–2023

    Postdoc at UChicago Data Science Institute

    He is a postdoctoral scholar at the University of Chicago's Data Science Institute from 2022 to 2023. His research there centers on reliability and robustness of statistical and machine learning methods, touching on generalization bounds, sequential decision making, learning theory and Markov chain Monte Carlo.

  6. July 2023

    A move to Waterloo

    He joins the University of Waterloo in July 2023, after his postdoc at the University of Chicago.

  7. 2025

    Thompson sampling for online learning

    In 2025, he and Alexander Terenin post a paper, Bayesian Algorithms for Adversarial Online Learning, that develops a form of Thompson sampling for online learning under full feedback. The paper shows that regret decomposes into the regret the learner expected a priori plus an excess regret term, and it recovers optimal rates in the finite-expert setting.

  8. As of October 2026

    Assistant professor at Waterloo

    He is an assistant professor in the Department of Statistics and Actuarial Science at the University of Waterloo and a faculty affiliate at the Vector Institute. His research interests are computational statistics, statistical learning theory and online learning.

Career

  1. NowUniversity of WaterlooAssistant Professor
  2. NowVector InstituteFaculty Affiliate
  3. 2022Simons Institute for the Theory of ComputingVisiting graduate student
  4. 2022–2023University of ChicagoPostdoctoral Scholar
  5. 2020Institute for Advanced StudyVisiting student

Education

  1. 2022University of TorontoMSc
  2. University of WaterlooBMath

From the biography.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Vector Institute. Our estimates of where the 943 people shaping AI who worked at Google and Google DeepMind work now: 446 at Google and 452 across 278 other organizations.
  2. Nonprofit roles are 2.3 times as common in AI safetyIncludes Vector Institute. We estimate where people focused on safety, alignment and evaluation work. Nonprofit roles appear for 25.1% of safety people with known affiliations, compared with 11.0% in other areas.
  3. Universities (31%) and startups and other companies (25%) employ the largest shares of 2020–2026 entrantsIncludes Vector Institute. Among 819 people whose first recorded role started in 2020–2026 and who have a known current organization, 60% work in commercial sectors (25% at startups and other companies, 23% at frontier labs and 13% at big tech, with 9 in two), and 31% work at universities, compared with 66% for earlier entrants.
  4. The United States employs an estimated 48% of people shaping AIIncludes Vector Institute. We estimate where the people shaping AI work today, how universities, companies and research institutes split across countries, and how country shares have changed since the 1960s.

Colleagues

People named in this biography, and people with dated positions at the same organization in the same years.

Sources and credits

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

  1. Information-Theoretic Generalization Bounds for SGLD via Data-Dependent Estimates arxiv.org. Checked 21 October 2026.
  2. In Defense of Uniform Convergence: Generalization via derandomization with an application to interpolating predictors arxiv.org. Checked 21 October 2026.
  3. Sharpened Generalization Bounds based on Conditional Mutual Information and an Application to Noisy, Iterative Algorithms arxiv.org. Checked 21 October 2026.
  4. Jeffrey Negrea | DSI datascience.uchicago.edu. Checked 21 October 2026.
  5. Minimax Optimal Quantile and Semi-Adversarial Regret via Root-Logarithmic Regularizers arxiv.org. Checked 21 October 2026.
  6. 2021 Rising Stars | DSI datascience.uchicago.edu. Checked 21 October 2026.
  7. Jeffrey Negrea - The Mathematics Genealogy Project mathgenealogy.org. Checked 21 October 2026.
  8. Jeffrey Negrea | Simons Institute for the Theory of Computing simons.berkeley.edu. Checked 21 October 2026.
  9. Jeffrey Negrea - Vector Institute for Artificial Intelligence vectorinstitute.ai. Checked 21 October 2026.
  10. Jeffrey Negrea | Math Innovation | University of Waterloo uwaterloo.ca. Checked 21 October 2026.
  11. Bayesian Algorithms for Adversarial Online Learning: from Finite to Infinite Action Spaces arxiv.org. Checked 21 October 2026.