AI People Wiki
Portrait of Nick Frosst
Photo: Gabriel Hutchinson, CC BY-SA 4.0. Credit

Builders & open source

Nick Frosst

Co-founder, Cohere

  • Co-founder, Cohere1

Nick Frosst is a co-founder of Cohere, according to the company's about page. Publishing as Nicholas Frosst, Frosst co-authored Dynamic Routing Between Capsules (2017) with Sara Sabour and Geoffrey Hinton and was first author of Analyzing and Improving Representations with the Soft Nearest Neighbor Loss (2019), both written with a Google Brain affiliation. Frosst is also a co-author of Neural Additive Models (2020), an approach to interpretable machine learning whose revised version lists Frosst with Cohere, and of Cohere's 2026 report on the North Small Translate model.12345

At a glance

Current role
Co-founder
Company
Cohere
Sector
Company
Work listed
3
Sources cited
8
Profile reviewed

Research at Google Brain

Publishing as Nicholas Frosst, Frosst worked with Geoffrey Hinton at Google Brain on capsule networks and interpretability. Frosst co-authored "Dynamic Routing Between Capsules" (2017) with Sara Sabour and Hinton, and with Hinton wrote "Distilling a Neural Network Into a Soft Decision Tree" (2017), which explains a trained network's decisions with a more interpretable tree model.26

Frosst was first author of DARCCC (2018), which detects adversarial examples by reconstructing inputs from class-conditional capsules, and of "Analyzing and Improving Representations with the Soft Nearest Neighbor Loss" (2019), written with Nicolas Papernot and Hinton. Frosst also co-authored Neural Additive Models (2020), an approach to interpretable machine learning with neural networks.734

Cohere

Frosst is a co-founder of Cohere, listed on the company's about page alongside co-founders Aidan Gomez and Ivan Zhang. Frosst appears among the authors of Cohere's technical reports, including the April 2025 report on Command A, an enterprise-focused large language model, and the September 2026 report on North Small Translate, a translation model.185

Work and contributions

  • Capsule networks and the soft nearest neighbor loss Published research

    Co-author, with Sara Sabour and Geoffrey Hinton, of Dynamic Routing Between Capsules (2017), and first author, with Nicolas Papernot and Geoffrey Hinton, of Analyzing and Improving Representations with the Soft Nearest Neighbor Loss (2019), both at Google Brain.23

  • Neural Additive Models Published research

    Co-author of Neural Additive Models: Interpretable Machine Learning with Neural Nets (2020), whose revised version lists Frosst with Cohere.4

  • North Small Translate Published research

    Co-author of Cohere's September 2026 report North Small Translate: Advanced Cost-Effective Translation (Cohere CAT+).5

Timeline

  1. Co-authored "Dynamic Routing Between Capsules" with Sara Sabour and Geoffrey Hinton.2
  2. Co-authored "Distilling a Neural Network Into a Soft Decision Tree" with Geoffrey Hinton.6
  3. First author of DARCCC, on detecting adversaries by reconstruction from class-conditional capsules.7
  4. First author of a paper analyzing and improving representations with the soft nearest neighbor loss.3
  5. Co-authored Neural Additive Models, an interpretable machine learning method.4
  6. Listed among the authors of Cohere's Command A technical report.8
  7. Co-authored Cohere's report on the North Small Translate model.5

Other documented roles

Roles the sources document that are not confirmed as current.

Where they work

Suggest a correction or removal

Profiles cover public professional work only, and each statement cites a source. If something is out of date, missing or wrong, tell us and link a source. People can also ask for their profile to be corrected or removed.

References

8 sources · profile reviewed Oct 21, 2026 · data updated Oct 21, 2026

  1. Cohere — About (leadership)Cohere · Official page · accessed Oct 21, 2026
  2. Dynamic Routing Between Capsules (arXiv 1710.09829), Google Brain affiliationarXiv · Paper · published Oct 26, 2017 · accessed Oct 21, 2026
  3. Analyzing and Improving Representations with the Soft Nearest Neighbor Loss (arXiv 1902.01889), Google Brain affiliationarXiv · Paper · published Feb 5, 2019 · accessed Oct 21, 2026
  4. Neural Additive Models: Interpretable Machine Learning with Neural Nets (arXiv 2004.13912), author affiliationsarXiv · Paper · published Apr 29, 2020 · accessed Oct 21, 2026
  5. North Small Translate: Advanced Cost-Effective Translation (Cohere CAT+) (arXiv 2609.13916)arXiv · Paper · published Sep 12, 2026 · accessed Oct 21, 2026
  6. Distilling a Neural Network Into a Soft Decision Tree (arXiv 1711.09784)arXiv · Paper · published Nov 27, 2017 · accessed Oct 21, 2026
  7. DARCCC: Detecting Adversaries by Reconstruction from Class Conditional Capsules (arXiv 1811.06969)arXiv · Paper · published Nov 16, 2018 · accessed Oct 21, 2026
  8. Command A: An Enterprise-Ready Large Language Model (arXiv 2504.00698)arXiv · Paper · published Apr 1, 2025 · accessed Oct 21, 2026