PeopleofSuperintelligence

Researcher

Anastasis Kratsios

Anastasis Kratsios is a mathematician whose work develops theory for deep learning, including universal feature maps, constrained transformers and neural operators. He earned a mathematics PhD at Concordia University in 2018, worked in ETH Zurich's Stochastic Finance Group from 2019 to 2021, and completed postdoctoral work with the SADA group before joining McMaster in 2022. His papers introduced NEU feature maps, proved approximation results for constrained transformers, embedded metric data with small transformers, and used attention-based neural operators for dynamic Stackelberg games. As of October 2026, he is an assistant professor at McMaster, a Vector Institute faculty affiliate, serves on the Fields Institute's Centre for Mathematical AI steering committee, and teaches its Mathematical Foundations of AI course.

Badges

Timeline

  1. 2018

    Mathematics PhD at Concordia

    He earns a mathematics PhD at Concordia University in 2018 with a thesis on arbitrage-free regularization, geometric learning, and non-Euclidean filtering in finance.

  2. 2019–2021

    Postdoctoral work at ETH Zurich

    From 2019 to 2021, he works as a postdoctoral researcher in ETH Zurich's Stochastic Finance Group.

  3. 2021

    NEU feature maps

    With Cody Hyndman, he introduces the Non-Euclidean Upgrading (NEU) meta-procedure in 2021. It learns feature maps that embed the universal approximation property in most model classes while ensuring the maps preserve that property.

  4. 2022

    A theorem for constrained transformers

    In 2022, he coauthors a theorem with Behnoosh Zamanlooy, Tianlin Liu and Ivan Dokmanić: probabilistic transformers uniformly approximate continuous functions while keeping randomized outputs inside a compact target set. The work also gives exact convex constraint satisfaction and a chart-free universal approximation theorem for Riemannian manifold-valued functions. The paper receives an ICLR 2022 Spotlight and a Top 3% designation.

  5. 2022

    From SADA to McMaster

    His postdoctoral work with the SADA group ends in 2022, and he joins McMaster University's Department of Mathematics and Statistics as an assistant professor in 2022.

  6. 2023

    Small transformers compute universal metric embeddings

    In 2023, he coauthors Small Transformers Compute Universal Metric Embeddings. The paper shows how probabilistic transformers embed any finite metric dataset into Gaussian-mixture representations with low metric distortion.

  7. 2025

    Neural operators in Stackelberg games

    In 2025, he coauthors a JMLR paper that uses an attention-based neural operator to approximate a follower's best-response operator uniformly across compact sets of adapted open-loop controls. Its game value approximates the value of the original Stackelberg game.

  8. As of October 2026

    McMaster professor on Fields AI committee

    As of October 2026, he is an assistant professor in McMaster University's Department of Mathematics and Statistics and a faculty affiliate at Vector Institute. He serves on the Steering Committee of the Fields Institute's Centre for Mathematical AI. In 2026, he teaches the Fields Institute's shared graduate course, Mathematical Foundations of AI.

Career

  1. Since 2026University of TorontoInstructor
  2. NowMcMaster UniversityAssistant Professor
  3. NowVector InstituteFaculty Affiliate
  4. Until 2022SADA groupPostdoctoral Researcher
  5. 2019–2021ETH ZurichPostdoctoral Researcher

Education

  1. 2018Concordia UniversityMathematics PhD

From the biography.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes University of Toronto. The median h-index is 48 at universities and 14 at frontier AI labs. We compare publishing careers, recent citations, dual appointments and job moves across 2,420 university researchers and 308 frontier-lab researchers.
  2. 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.
  3. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes University of Toronto. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  4. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorateIncludes University of Toronto. 84% of people shaping AI with recorded undergraduate and doctoral schools changed institutions for their doctorate. Across 2,116 people with both an undergraduate degree and a doctorate recorded, we trace which undergraduate schools feed which doctoral programs, how often students stay at the same school, and how paths differ by birth-country region.

Colleagues

People with dated positions at the same organization in the same years.

Sources and credits

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

  1. Arbitrage-free regularization, geometric learning, and non-Euclidean filtering in finance spectrum.library.concordia.ca. Checked 21 October 2026.
  2. josef teichmann :: personal homepage at ETH Zurich people.math.ethz.ch. Checked 21 October 2026.
  3. NEU: A Meta-Algorithm for Universal UAP-Invariant Feature Representation jmlr.org. Checked 21 October 2026.
  4. Universal Approximation Under Constraints Is Possible with Transformers experts.mcmaster.ca. Checked 21 October 2026.
  5. sada-lab.ch | Anastasis Kratsios (PostDoc '22) sada.dmi.unibas.ch. Checked 21 October 2026.
  6. Anastasis Kratsios experts.mcmaster.ca. Checked 21 October 2026.
  7. Small Transformers Compute Universal Metric Embeddings jmlr.org. Checked 21 October 2026.
  8. Neural Operators Can Play Dynamic Stackelberg Games jmlr.org. Checked 21 October 2026.
  9. Anastasis Kratsios - Vector Institute for Artificial Intelligence vectorinstitute.ai. Checked 21 October 2026.
  10. Centre for Mathematical AI | Fields Institute for Research in Mathematical Sciences fieldsinstitute.ca. Checked 21 October 2026.
  11. Fields Academy Shared Graduate Course: Mathematical Foundations of AI www2.fields.utoronto.ca. Checked 21 October 2026.