PeopleofSuperintelligence

Researcher

Darius Dabert

As of October 2026, Darius Dabert is an AI researcher whose coauthored work spans GPU-accelerated neural networks, constraint programming, and Mistral model research. He worked as a research intern at Montréal's CORAIL group in 2024 and coauthored work on fused MLPs for Intel GPUs. His 2025 papers covered learned bounds for constraint programming, Magistral reasoning, Voxtral audio chat, and Devstral code agents. His 2026 work covered compact language models, streaming speech, text-to-speech, multimodal safety, and robot navigation.

Works
8
Citations
8
h-index
2

Publication numbers from OpenAlex, as of 19 October 2026.

Timeline

  1. 2024

    Research at CORAIL

    He works as a CORAIL research intern in 2024 and receives an Ecole Polytechnique research internship award.

  2. 2024

    Fused MLPs for Intel GPUs

    In 2024, he coauthors a paper on a SYCL implementation of multilayer perceptrons optimized for Intel's Data Center GPU Max 1550. The Intel tiny-dpcpp-nn project credits him with its original SwiftNet implementation.

  3. 2025

    Learning dual bounds

    At Polytechnique Montréal and École Polytechnique, he coauthors a 2025 AAAI paper that uses self-supervised neural networks to learn valid dual bounds for constraint programming.

  4. 2025

    Mistral's first reasoning model

    In 2025, he coauthors Magistral, Mistral AI's first reasoning model, and a report on its scalable reinforcement-learning pipeline.

  5. 2025

    Audio chat and code agents

    In 2025, he coauthors Voxtral, a multimodal audio chat model, and Devstral-Small, a lightweight open-source model for code agents.

  6. 2026

    The Ministral 3 family

    In 2026, he coauthors the Ministral 3 report, which introduces parameter-efficient dense language models in 3B, 8B, and 14B sizes.

  7. 2026

    Streaming and synthetic speech

    In 2026, he coauthors Voxtral Realtime, a streaming speech-recognition model, and Voxtral TTS, a multilingual text-to-speech model.

  8. As of October 2026

    Safety and robot navigation

    He is an AI researcher coauthoring 2026 work on Shieldstral, a multimodal safety classifier, and Robostral Navigate, an 8B vision-language model for robot navigation.

Most-cited works

  1. Ministral 32026 · 4 citationsWith Abhinav Rastogi, Alexander H. Liu, Alexandre Sablayrolles, Amélie Héliou, Andy Ehrenberg and Andy Lo
  2. Magistral2025 · 3 citationsWith Alexander H. Liu, Alexandre Sablayrolles, Amélie Héliou, Andy Ehrenberg, Baptiste Bout and Baudouin De Monicault
  3. Fully-fused Multi-Layer Perceptrons on Intel Data Center GPUs2024 · 1 citations
  4. Robostral Navigate2026 · 0 citationsWith Alexandre Sablayrolles, Amélie Héliou, Baptiste Bout, Baudouin De Monicault, Chris Bamford and Christian Wallenwein
  5. Shieldstral2026 · 0 citationsWith Alexandre Sablayrolles, Amélie Héliou, Anmol Agarwal, Baptiste Bout, Baudouin De Monicault and Chris Bamford

From OpenAlex, as of 19 October 2026.

  1. Google had the most people shaping AI every year from 2014 to 2025Our estimates of which organizations employed the most people shaping AI from 1961 to 2025, with a year-by-year chart.
  2. The United States employs an estimated 53% of people in our data with a known employment countryOur estimates of where AI careers are concentrated now and how countries’ shares of dated positions changed over time.
  3. 47% of living people in our data with known birth and employment countries work outside their birth countryOur estimate of cross-border careers in AI, from country of birth and the country of each current organization.
  4. More people shaping AI studied at Massachusetts Institute of Technology than at any other institutionOur estimates of where the people shaping AI studied and where their careers led, from education records in cited biographies.

Sources and credits

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

  1. Corail Research Group - Team corail-research.github.io. Checked 19 October 2026.
  2. Fully-fused Multi-Layer Perceptrons on Intel Data Center GPUs arxiv.org. Checked 19 October 2026.
  3. GitHub - intel/tiny-dpcpp-nn: SYCL implementation of Fused MLPs for Intel GPUs github.com. Checked 19 October 2026.
  4. Learning Valid Dual Bounds in Constraint Programming: Boosted Lagrangian Decomposition with Self-Supervised Learning | Proceedings of the AAAI Conference on Artificial Intelligence ojs.aaai.org. Checked 19 October 2026.
  5. Magistral arxiv.org. Checked 19 October 2026.
  6. Voxtral arxiv.org. Checked 19 October 2026.
  7. Devstral: Fine-tuning Language Models for Coding Agent Applications arxiv.org. Checked 19 October 2026.
  8. Ministral 3 arxiv.org. Checked 19 October 2026.
  9. Voxtral Realtime arxiv.org. Checked 19 October 2026.
  10. Voxtral TTS arxiv.org. Checked 19 October 2026.
  11. Shieldstral arxiv.org. Checked 19 October 2026.
  12. Robostral Navigate arxiv.org. Checked 19 October 2026.