PeopleofSuperintelligence

Researcher

Emilien Dupont

Emilien Dupont is a machine-learning researcher whose work spans generative models, neural differential equations and AI-assisted algorithm discovery. He studied theoretical physics at Imperial College London and computational mathematics at Stanford, then developed deep generative models for interpretable representations at Schlumberger. He introduced Joint-VAE and Augmented Neural ODEs before completing a machine-learning PhD at Oxford in 2022 and presenting functa, a framework for treating neural functions as data. He is a senior research scientist at Google DeepMind, where he helps develop AlphaEvolve and coauthors research that improves an upper bound for matrix multiplication.

Links
Website, Google Scholar, X and GitHub

Timeline

  1. 2010–2018

    Physics degrees and Schlumberger

    He studies theoretical physics at Imperial College London from 2010 to 2014 and completes an MS in computational mathematics at Stanford in 2016. From 2016 to 2018, he works as a machine-learning scientist at Schlumberger, developing algorithms for time series, vision and geology and researching interpretable deep generative models.

  2. 2018

    Joint-VAE representations

    In 2018, he publishes Joint-VAE, a framework that learns disentangled continuous and discrete representations without supervision.

  3. 2019

    Augmented Neural ODEs

    In 2019, with Arnaud Doucet and Yee Whye Teh, he introduces Augmented Neural ODEs, which expand representational capacity and improve stability, generalization and computational cost.

  4. 2022

    Oxford PhD and functa

    He completes a machine-learning PhD at Oxford in 2022 with Yee Whye Teh and Arnaud Doucet. With Hyunjik Kim and collaborators, he presents functa, a framework for deep learning on neural functions treated as data.

  5. Since November 2024

    Senior scientist at Google DeepMind

    He is a senior research scientist at Google DeepMind since November 2024, after joining as a research scientist in January 2023 to work on LLMs for scientific discovery and neural compression. In May 2025, he helps develop AlphaEvolve, an evolutionary coding agent that combines Gemini models with automated evaluators to discover algorithms; by May 2026, the system reaches applications in physics, electricity grids and computing infrastructure. In August 2026, he coauthors research using AlphaEvolve to improve the upper bound on the matrix-multiplication exponent to ω < 2.371177.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 46% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

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 Emilien Dupont yet. Send us a free-licensed photo.

  1. Resume | Emilien Dupont emiliendupont.github.io. Checked 19 October 2026.
  2. OxCSML : People : Emilien Dupont csml.stats.ox.ac.uk. Checked 19 October 2026.
  3. Learning Disentangled Joint Continuous and Discrete Representations arxiv.org. Checked 19 October 2026.
  4. Augmented Neural ODEs papers.nips.cc. Checked 19 October 2026.
  5. From data to functa: Your data point is a function and you can treat it like one proceedings.mlr.press. Checked 19 October 2026.
  6. AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms — Google DeepMind deepmind.google. Checked 19 October 2026.
  7. AlphaEvolve: Gemini-powered coding agent scaling impact across fields — Google DeepMind deepmind.google. Checked 19 October 2026.
  8. Improving the matrix multiplication exponent with modern optimization and AlphaEvolve arxiv.org. Checked 19 October 2026.