PeopleofSuperintelligence

Researcher

Mohammadamin Barekatain

Also known as Amin Barekatain

Mohammadamin Barekatain is a quantitative developer and former DeepMind research engineer who helped build AlphaTensor, AlphaDev and FunSearch, systems that use reinforcement learning and language models to discover new algorithms and mathematical constructions. He studied computer engineering at the University of Isfahan and computer science at the Technical University of Munich, and in 2017 he was first author of the Okutama-Action aerial video dataset. After co-authoring the Neural A* path-planning method at OMRON SINIC X in Tokyo in 2020, he joined DeepMind in London and became a senior research engineer in 2022. The 2022 AlphaTensor paper on matrix multiplication and the 2023 Nature papers on AlphaDev and FunSearch list him among their authors. Since January 2024, he is a quantitative developer team lead at Quadrature in London.

Now at
Quadrature
Works
16
Citations
2,625
h-index
12
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2011–2019

    Isfahan and Munich degrees

    He studies computer engineering at the University of Isfahan from 2011 to 2015 and earns a BA. From 2015 to 2019 he studies computer science at the Technical University of Munich and earns an MSc.

  2. 2017

    Okutama-Action aerial dataset

    As a Munich student in the National Institute of Informatics International Internship Program in Tokyo, he is first author of Okutama-Action, a video dataset for aerial-view human action detection. The 2017 CVPR Workshops paper describes 43 minute-long, fully annotated sequences with 12 action classes.

  3. 2020

    Neural A* path planning

    In 2020, he co-authors Neural A* with Ryo Yonetani, Tatsunori Taniai, Mai Nishimura and Asako Kanezaki while at OMRON SINIC X in Tokyo. The method turns the A* search algorithm into a differentiable module coupled with a convolutional encoder, forming an end-to-end trainable path planner that is accepted to ICML 2021.

  4. 2020–2024

    Research engineer at DeepMind

    He joins DeepMind in London in 2020 and is a senior research engineer there from November 2022 until January 2024. In 2021, he is an author of MuZero Unplugged, a single unified reinforcement-learning algorithm for any data budget, including offline RL.

  5. October 2022

    AlphaTensor for matrix multiplication

    In October 2022, he is among the DeepMind authors of the Nature paper on AlphaTensor, an agent trained to play a single-player game whose goal is finding tensor decompositions. AlphaTensor discovers matrix-multiplication algorithms that outperform the state-of-the-art complexity for many matrix sizes.

  6. June 2023

    AlphaDev sorting routines

    In June 2023, he is among the DeepMind authors of the Nature paper on AlphaDev, a deep reinforcement learning agent. AlphaDev discovers small sorting algorithms from scratch that outperform previously known human benchmarks.

  7. December 2023

    FunSearch and cap sets

    In December 2023, he is one of eight equal-contribution authors of the Nature paper on FunSearch, which pairs a pretrained large language model with a systematic evaluator. FunSearch finds new constructions of large cap sets in extremal combinatorics and improved heuristics for online bin packing.

  8. Since January 2024

    Quantitative developer at Quadrature

    In January 2024, he leaves DeepMind for Quadrature in London, where he is a quantitative developer team lead.

Career

  1. Since 2024QuadratureQuantitative Developer Team Lead
  2. 2020–2024Google DeepMindResearch Engineer, Senior Research Engineer

Education

  1. 2019Technical University of MunichMSc
  2. 2015University of IsfahanBA

From the biography and OpenAlex affiliations (the years of papers that list the organization), so some years are approximate.

Most-cited works

  1. Machine Learning for Internet of Things Data Analysis: A Survey2017 · 1,002 citationsWith Amit Sheth
  2. Mathematical discoveries from program search with large language models2023 · Nature · 522 citationsWith Alexander Novikov, Bernardino Romera-Paredes, Emilien Dupont, Francisco J. R. Ruiz, Matej Balog and Pushmeet Kohli
  3. Discovering faster matrix multiplication algorithms with reinforcement learning2022 · Nature · 478 citationsWith Aja Huang, Alexander Novikov, Bernardino Romera-Paredes, David Silver, Demis Hassabis and Francisco J. R. Ruiz
  4. Okutama-Action: An Aerial View Video Dataset for Concurrent Human Action Detection2017 · IEEE Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · 204 citations
  5. Faster sorting algorithms discovered using deep reinforcement learning2023 · Nature · 176 citationsWith David Silver, Demis Hassabis, Julian Schrittwieser, Martin Riedmiller, Oriol Vinyals and Pushmeet Kohli

From OpenAlex, as of 21 October 2026.

  1. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Mohammadamin Barekatain. 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. 16% of AI career moves in the 2020s went from a university to a companyIncludes Mohammadamin Barekatain. We estimate that 16% of AI career moves in the 2020s went from a university to a company; 29% of moves starting at a university went to a company, and 76% ended the university role.
  3. Google had the most people shaping AI every year from 2014 to 2025Includes Google. Our estimates of which organizations employed the most people shaping AI each year from 1955 to 2025, tracing the shift from university labs to companies.
  4. Google has the highest combined h-index among AI organizationsIncludes Google. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.

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 Mohammadamin Barekatain yet. Send us a free-licensed photo.

  1. Mohammadamin Barekatain | alphaXiv alphaxiv.org. Checked 20 October 2026.
  2. Okutama-Action: An Aerial View Video Dataset for Concurrent Human Action Detection (arXiv 1706.03038) ar5iv.labs.arxiv.org. Checked 20 October 2026.
  3. Okutama-Action: An Aerial View Video Dataset for Concurrent Human Action Detection (CVPR 2017 Workshops) openaccess.thecvf.com. Checked 20 October 2026.
  4. [2009.07476] Path Planning using Neural A* Search arxiv.org. Checked 20 October 2026.
  5. Path Planning using Neural A* Search (arXiv HTML) arxiv.org. Checked 20 October 2026.
  6. [2104.06294] Online and Offline Reinforcement Learning by Planning with a Learned Model arxiv.org. Checked 20 October 2026.
  7. Discovering faster matrix multiplication algorithms with reinforcement learning nature.com. Checked 20 October 2026.
  8. Discovering novel algorithms with AlphaTensor — Google DeepMind deepmind.google. Checked 20 October 2026.
  9. Faster sorting algorithms discovered using deep reinforcement learning nature.com. Checked 20 October 2026.
  10. Mathematical discoveries from program search with large language models nature.com. Checked 20 October 2026.
  11. Mohammadamin Barekatain - Quantitative Developer Team Lead at Quadrature | The Org theorg.com. Checked 20 October 2026.