PeopleofSuperintelligence

Researcher

Laurent Sartran

Laurent Sartran is a research engineer who co-led AlphaProof, the Google DeepMind system that, together with AlphaGeometry 2, reached silver-medal level at the 2024 International Mathematical Olympiad. After working as a software engineer at Cisco from 2014 to 2017, they joined DeepMind as a research engineer in 2018 and became a senior research engineer in October 2020. Their language-model work included Transformer Grammars, a 2022 paper that added recursive syntactic composition to Transformers, and CDCD, a 2022 framework for applying continuous diffusion to categorical data such as language. With Thomas Hubert and Rishi Mehta, Sartran provided the primary scientific, technical and strategic leadership of AlphaProof, which Nature published in November 2025. They are a member of technical staff at Anthropic, which they joined in 2025.

Now at
Anthropic

Timeline

  1. 2014–2017

    Software engineer at Cisco

    Sartran works as a software engineer at Cisco from 2014 to 2017. In May 2016, Cisco files a patent naming them as a co-inventor of edge-based detection of new and unexpected network flows, in which a device builds an anomaly detection model for new interactions between network nodes.

  2. 2018–2020

    Research engineer at DeepMind

    They join DeepMind as a research engineer in 2018. In 2020, they coauthor a TACL paper with Lei Yu, Chris Dyer and others showing that Bayes' rule yields document translation models that can learn from parallel sentences and monolingual documents.

  3. October 2020–2025

    Tech lead on AlphaProof

    In October 2020, Sartran becomes a senior research engineer at Google DeepMind, a role held until 2025. They serve as tech lead on AlphaProof and contribute to the DeepMind JAX Ecosystem, a collection of libraries for machine learning research.

  4. 2022

    Transformer Grammars

    As first author, Sartran publishes Transformer Grammars in TACL in 2022 with Samuel Barrett, Adhiguna Kuncoro, Miloš Stanojević, Phil Blunsom and Chris Dyer. The models combine Transformers with recursive syntactic composition, implemented through a special attention mask, and outperform strong baselines on sentence-level perplexity and syntax-sensitive evaluations.

  5. 2022–2023

    Diffusion for text and SynJax

    In November 2022, they coauthor CDCD with Sander Dieleman and others, a framework for modelling categorical data such as language with diffusion models that are continuous in time and input space. In August 2023, they and Miloš Stanojević present SynJax, a JAX library of vectorized inference algorithms for structured distributions such as constituency trees and spanning trees.

  6. July 2024

    AlphaProof at the 2024 IMO

    Thomas Hubert, Rishi Mehta and Sartran lead the development of AlphaProof, a reinforcement-learning system for formal math reasoning. In July 2024, Google DeepMind announces that AlphaProof and AlphaGeometry 2 together solve four of six problems from the 2024 International Mathematical Olympiad, the same level as a silver medalist.

  7. November 2025

    The AlphaProof paper in Nature

    Nature publishes the AlphaProof paper in November 2025, with Sartran listed third after Thomas Hubert and Rishi Mehta. The three provide the project's primary scientific, technical and strategic leadership, and Sartran leads the Lean environment development with Eric Wieser.

  8. Since 2025

    Member of technical staff at Anthropic

    Sartran leaves Google DeepMind and joins Anthropic as a member of technical staff in 2025.

Career

  1. Since 2025AnthropicMember of Technical Staff
  2. 2018–2025Google DeepMindResearch Engineer, Senior Research Engineer
  3. 2014–2017Cisco SystemsSoftware Engineer

From the biography.

Videos

YouTube videos in which Laurent Sartran speaks or appears, newest first. A model chose them from YouTube search results. List updated 20 October 2026.

  1. No Priors Ep. 90 | With Google's DeepMind's AlphaProof TeamNo Priors: AI, Machine Learning, Tech, & Startups · 14 November 2024 · 39 min

Video data from YouTube API Services. By watching you agree to the YouTube Terms of Service. How we use YouTube data.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Laurent Sartran. 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. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Laurent Sartran. We estimate that alumni of 17 major AI labs went on to found or lead 350 other organizations, including 226 they founded, and we trace 472 career moves directly between 23 labs since 2010.
  3. 22 of 133 people shaping AI with a former OpenAI role now work at AnthropicIncludes Anthropic. Our estimates of where OpenAI recruits came from and where 133 people with former OpenAI roles work now, including the companies and labs they founded.
  4. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Anthropic. 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.

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 Laurent Sartran yet. Send us a free-licensed photo.

  1. Laurent Sartran | alphaXiv alphaxiv.org. Checked 20 October 2026.
  2. US10389741B2 - Edge-based detection of new and unexpected flows - Google Patents patents.google.com. Checked 20 October 2026.
  3. Better Document-Level Machine Translation with Bayes’ Rule - ACL Anthology aclanthology.org. Checked 20 October 2026.
  4. Transformer Grammars: Augmenting Transformer Language Models with Syntactic Inductive Biases at Scale - ACL Anthology aclanthology.org. Checked 20 October 2026.
  5. [2211.15089] Continuous diffusion for categorical data arxiv.org. Checked 20 October 2026.
  6. [2308.03291] SynJax: Structured Probability Distributions for JAX arxiv.org. Checked 20 October 2026.
  7. AI achieves silver-medal standard solving International Mathematical Olympiad problems — Google DeepMind deepmind.google. Checked 20 October 2026.
  8. Olympiad-level formal mathematical reasoning with reinforcement learning | Nature nature.com. Checked 20 October 2026.
  9. Olympiad-Level Formal Mathematical Reasoning with Reinforcement Learning research.google. Checked 20 October 2026.
  10. lsartran (Laurent Sartran) · GitHub github.com. Checked 20 October 2026.