PeopleofSuperintelligence

Researcher

Edouard Grave

Also known as E. Grave

Edouard Grave is an AI researcher whose work spans distributional semantics, multilingual word vectors, foundation language models and speech systems. He earned a computer-science PhD at Université Paris VI in 2014; his training also includes an École Polytechnique MSc and postdoctoral work at UC Berkeley and Columbia. At Facebook AI Research, he co-led fastText and LLaMA before joining Kyutai’s scientific team, where he coauthored Moshi and Hibiki. By October 2026, he is a Kyutai alumnus who held the title Chief Language Officer.

Now at
Kyutai
Born
1986
Badges

Years link to their sources.

Timeline

  1. 2014

    A PhD in distributional semantics

    He defends his PhD in computer science at Université Paris VI in January 2014, with Francis Bach and Guillaume Obozinski as supervisors. His training includes an MSc in machine learning and computer vision from École Polytechnique and postdoctoral work at UC Berkeley and Columbia University.

  2. 2016–2017

    The fastText papers

    He coauthors the 2016 paper Enriching Word Vectors with Subword Information, which represents words as bags of character n-grams, and the 2017 fastText classifier paper. His work at Facebook AI Research includes co-leading development of the fastText library; its classifier is often on par with deep-learning classifiers in accuracy and many orders of magnitude faster in training and evaluation.

  3. 2018

    Word vectors for 157 languages

    In 2018, he coauthors a method for training high-quality word vectors in 157 languages from Wikipedia and Common Crawl, with new word-analogy datasets for French, Hindi and Polish.

  4. 2023

    LLaMA at FAIR

    At Facebook AI Research, he co-leads development of LLaMA. In 2023, he coauthors its paper with 13 other researchers; it introduces 7B-to-65B-parameter foundation models trained on trillions of tokens.

  5. 2023–2024

    Joining Kyutai’s scientific team

    In November 2023, he joins Kyutai’s scientific team as the Paris lab launches its open AI research program. In 2024, he coauthors Moshi, a speech-text foundation model for real-time dialogue that casts conversation as speech-to-speech generation and models the user’s and system’s speech in parallel streams.

  6. 2025

    Hibiki for speech translation

    In 2025, he coauthors Hibiki, a decoder-only model for simultaneous speech translation. It processes source and target speech through a multistream language model and produces text and audio tokens for speech translation.

  7. By 2026

    Compressing context for LLMs

    In 2026, he coauthors ARC-Encoder, which compresses context into continuous representations that replace token embeddings in decoder language models. By October 2026, he is a Kyutai alumnus who held the title Chief Language Officer.

Career

  1. Until 2023FAIR (Meta)
  2. From 2023KyutaiChief Language Officer
  3. University of California, BerkeleyPostdoctoral researcher
  4. Columbia UniversityPostdoctoral researcher

Education

  1. 2014Sorbonne UniversityPhD
  2. École PolytechniqueMSc

From the biography.

  1. 16% of AI career moves in the 2020s went from a university to a companyIncludes Edouard Grave. 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.
  2. Google alumni have founded or led 133 other organizations, more than any other AI labIncludes Kyutai. 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. 19% of people shaping AI who founded a dated organization have founded two or more, waiting a median of 7 years before a secondIncludes Kyutai. Across 1,136 people shaping AI who founded or co-founded a dated company, standalone lab or nonprofit, 221 (19%) founded at least two organizations, accounting for 36% of all 1,421 founding events.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionIncludes University of California, Berkeley. We estimate where the people shaping AI studied and where their careers led. Massachusetts Institute of Technology leads with 467 former students, including 291 recorded doctorates.

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 Edouard Grave yet. Send us a free-licensed photo.

  1. ALPAGE - 2014 - Rapport annuel d'activité radar.inria.fr. Checked 21 October 2026.
  2. Launch of Kyutai – Europe’s first independent research lab dedicated to AI open science, co-founded by the iliad Group, CMA CGM and Schmidt Futures cmacgm-group.com. Checked 21 October 2026.
  3. Enriching Word Vectors with Subword Information arxiv.org. Checked 21 October 2026.
  4. Bag of Tricks for Efficient Text Classification aclanthology.org. Checked 21 October 2026.
  5. Learning Word Vectors for 157 Languages arxiv.org. Checked 21 October 2026.
  6. LLaMA: Open and Efficient Foundation Language Models arxiv.org. Checked 21 October 2026.
  7. Papers kyutai.org. Checked 21 October 2026.
  8. Kyutai Team kyutai.org. Checked 21 October 2026.