PeopleofSuperintelligence

Researcher

Pol Moreno

Also known as Pol Moreno Comellas

Pol Moreno is an AI researcher whose work spans computer vision, multi-agent learning, sports analytics and mathematical reasoning. His 2013 MSc project studies dependencies in neural recordings, and he presents inverse-graphics research at ECCV in 2016. At Edinburgh, he is a PhD student under Chris Williams and interns at DeepMind from August to December 2017, developing generative methods for scene descriptions. He contributes to Google DeepMind's 2021 football-analytics overview and coauthors a paper on recursive belief states; in 2024, he coauthors TacticAI, a Liverpool FC collaboration for corner-kick analysis. In 2025, he is named among the IMO team's leads and key contributors as Gemini Deep Think reaches gold-medal level, and as of October 2026 he is part of the Deep Think team behind a Gemini-backed Paper Assistant tested at STOC 2026.

Now at
Google
Works
11
Citations
359
h-index
8
Badges

Publication numbers from OpenAlex, as of 22 October 2026.

Timeline

  1. 2013

    MSc project under Iain Murray

    His MSc project under Iain Murray studies machine learning of dependencies in high-density neural recordings in 2013.

  2. 2016

    Overcoming occlusion with inverse graphics

    In 2016, he coauthors Overcoming Occlusion with Inverse Graphics with Christopher K. I. Williams, Charlie Nash and Pushmeet Kohli, and presents it orally at ECCV’s Geometry Meets Deep Learning workshop.

  3. 2017

    PhD and DeepMind internship

    In 2017, he is a PhD student at the University of Edinburgh under Chris Williams, with research interests in machine learning and computer vision. From August to December 2017, he works as a DeepMind research intern; his research combines generative models and deep learning to produce interpretable, accurate scene descriptions.

  4. 2021

    AI for football analytics

    In 2021, he contributes to Google DeepMind’s overview of sports analytics. It presents sports as a testbed for AI in complex multiagent environments and connects football analysis to computer vision, statistical learning and game theory.

  5. 2021

    Recursive belief states

    In 2021, he coauthors Neural Recursive Belief States in Multi-Agent Reinforcement Learning. The paper uses recursive deep generative models to approximate hierarchical beliefs about other agents, and its belief-model agents outperform model-free baselines with equivalent representational capacity.

  6. 2024

    TacticAI for corner kicks

    He coauthors TacticAI in 2024, a football tactics assistant developed and evaluated with Liverpool FC experts for corner-kick analysis. Its predictive and generative components let coaches explore player setups and recommend position adjustments.

  7. 2025

    Gemini Deep Think at IMO

    He is named among the leads and key contributors to the IMO 2025 team. Gemini Deep Think solves five of six problems perfectly, earns 35 points and reaches gold-medal-level performance.

  8. As of October 2026

    Deep Think and Paper Assistant

    As of October 2026, he is part of Google’s Deep Think team. The team contributes to Paper Assistant Tool, a Gemini 2.5 Deep Think system tested for STOC 2026 to give authors automated feedback on conference papers.

Career

  1. Since 2021Google
  2. 2021–2024Google DeepMindResearch Intern

Education

  1. University of EdinburghMSc

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

Most-cited works

  1. Game Plan: What AI can do for Football, and What Football can do for AI2021 · Journal of Artificial Intelligence Research · 116 citationsWith Alex Bridgland, Demis Hassabis, Michal Valko and Thore Graepel
  2. TacticAI: an AI assistant for football tactics2024 · Nature Communications · 90 citationsWith Demis Hassabis and Michael Bowling
  3. NeRF-VAE: A Geometry Aware 3D Scene Generative Model2021 · 45 citationsWith Danilo Jimenez Rezende
  4. Vision-as-Inverse-Graphics: Obtaining a Rich 3D Explanation of a Scene from a Single Image2017 · IEEE International Conference on Computer Vision Workshops · 30 citationsWith Pushmeet Kohli
  5. Overcoming Occlusion with Inverse Graphics2016 · Lecture notes in computer science · 28 citationsWith Pushmeet Kohli

From OpenAlex, as of 21 October 2026.

  1. Of 943 people shaping AI who worked at Google, 446 work there nowIncludes Pol Moreno. 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. 16% of AI career moves in the 2020s went from a university to a companyIncludes Pol Moreno. 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. The median h-index is 48 at universities and 14 at frontier AI labsIncludes Pol Moreno. 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.
  4. 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.

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 Pol Moreno yet. Send us a free-licensed photo.

  1. Projects supervised by Iain Murray homepages.inf.ed.ac.uk. Checked 21 October 2026.
  2. PhD Studentship in Vision as Inverse Graphics homepages.inf.ed.ac.uk. Checked 21 October 2026.
  3. Pol Moreno Comellas polmorenoc.github.io. Checked 21 October 2026.
  4. Advancing sports analytics through AI research deepmind.google. Checked 21 October 2026.
  5. Neural Recursive Belief States in Multi-Agent Reinforcement Learning arxiv.org. Checked 21 October 2026.
  6. TacticAI: an AI assistant for football tactics nature.com. Checked 21 October 2026.
  7. Advanced version of Gemini with Deep Think officially achieves gold-medal standard at the International Mathematical Olympiad deepmind.google. Checked 21 October 2026.
  8. Gemini-backed Paper Assistant Tool provides automated feedback for theoretical computer scientists at STOC 2026 research.google. Checked 21 October 2026.