PeopleofSuperintelligence

Researcher

Anna Potapenko

Anna Potapenko is an author on two AlphaFold papers in Nature from DeepMind, one on protein structure prediction and one on the joint structure of complexes. Her earlier work includes a 2013 ECIR paper with Konstantin Vorontsov, 'Robust PLSA Performs Better Than LDA', on topic modeling. In 2019 she co-authored the Compressive Transformer, a long-range sequence model, and in 2020 she co-authored an ACL paper on multi-agent communication and natural language learning. The AlphaFold paper of July 2021 lists her among the DeepMind authors. Her most recent dated affiliation is Google DeepMind in London, as listed on the 2024 AlphaFold 3 paper.

Timeline

  1. 2013

    Robust PLSA at ECIR 2013

    With Konstantin Vorontsov, she publishes 'Robust PLSA Performs Better Than LDA', a topic-model paper, in the ECIR 2013 proceedings.

  2. 2019

    The Compressive Transformer

    In November 2019 she co-authors the Compressive Transformer with Jack W. Rae, Siddhant M. Jayakumar and Timothy P. Lillicrap, a long-range sequence model.

  3. 2020

    Multi-agent language learning

    With Angeliki Lazaridou and Olivier Tieleman, she co-authors a method for combining multi-agent communication with natural language learning, published at ACL 2020.

  4. July 2021

    AlphaFold in Nature

    She is one of the DeepMind authors of AlphaFold's paper in Nature, published in July 2021, which predicts protein structures with atomic accuracy.

  5. May 2024

    AlphaFold 3 for complexes

    She is among the Google DeepMind authors of AlphaFold 3, which Nature publishes in May 2024, describing a diffusion-based architecture for the joint structure of complexes.

  6. Since 2024

    Google DeepMind in London

    Her most recent dated affiliation is Google DeepMind in London, as listed on the 2024 AlphaFold 3 paper.

Sources and credits

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

  1. Robust PLSA Performs Better Than LDA publications.hse.ru. Checked 14 October 2026.
  2. [1911.05507] Compressive Transformers for Long-Range Sequence Modelling arxiv.org. Checked 14 October 2026.
  3. Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning - ACL Anthology aclanthology.org. Checked 14 October 2026.
  4. [2005.07064] Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning arxiv.org. Checked 14 October 2026.
  5. Highly accurate protein structure prediction with AlphaFold | Nature nature.com. Checked 14 October 2026.
  6. Accurate structure prediction of biomolecular interactions with AlphaFold 3 | Nature nature.com. Checked 14 October 2026.