PeopleofSuperintelligence

Researcher

Hadrien Hendrikx

Hadrien Hendrikx is a machine learning researcher who works on distributed optimization and, more recently, on machine learning for ecology. He graduated from Ecole Polytechnique in 2016, took a master's degree at EPFL in 2018, and did his PhD at Inria Paris from 2018 to 2021 under Francis Bach and Laurent Massoulié. A 2021 paper he coauthored on continuized Nesterov acceleration and gossip algorithms won a NeurIPS Outstanding Paper Award. After a postdoc in the MLO team at EPFL from October 2021 to December 2022, he coauthored a 2025 paper on Byzantine-robust gossip and a 2026 preprint on soil fauna image analysis. He is a researcher in the Thoth team at Inria Grenoble, where he co-leads the XLBiod chair on explainable machine learning for ecology.

Now at
Inria
Badges

Timeline

  1. 2016–2018

    Polytechnique, then an EPFL master

    He graduates from Ecole Polytechnique in 2016 and earns a master's degree in computer science at EPFL in 2018. During the master's, he works as a research assistant in Rachid Guerraoui's DCL lab and coauthors a paper that defines dynamic safe interruptibility for decentralized multi-agent reinforcement learning.

  2. 2018–2021

    PhD with Bach and Massoulié

    From 2018 to 2021, he is a PhD student in the SIERRA and DYOGENE teams at Inria Paris, which are part of the computer science department of Ecole Normale Supérieure. He works on decentralized optimization under Francis Bach and Laurent Massoulié.

  3. 2020

    Preconditioned accelerated gradient method

    In 2020, he coauthors an International Conference on Machine Learning paper with Lin Xiao, Sébastien Bubeck, Francis Bach and Laurent Massoulié on distributed empirical risk minimization. The paper proposes a preconditioned accelerated gradient method, designed to reduce the number of communications, in which the server solves a local optimization problem over a subsampled dataset.

  4. 2021

    A NeurIPS Outstanding Paper Award

    In 2021, "Continuized Accelerations of Deterministic and Stochastic Gradient Descents, and of Gossip Algorithms", which he writes with Mathieu Even, Raphaël Berthier, Francis Bach, Nicolas Flammarion, Pierre Gaillard, Laurent Massoulié and Adrien Taylor, wins a NeurIPS Outstanding Paper Award. The paper introduces a continuized version of Nesterov acceleration with gradient steps at random times and gives the first rigorous acceleration of asynchronous gossip algorithms.

  5. October 2021–December 2022

    Postdoc in EPFL's MLO team

    From October 2021 to December 2022, he is a postdoc in the MLO team at EPFL, working with Martin Jaggi. With Thijs Vogels and Jaggi, he publishes "Beyond spectral gap: the role of the topology in decentralized learning" in the NeurIPS 2022 main conference track, a paper that quantifies how the graph topology influences convergence.

  6. 2025

    Byzantine-robust gossip algorithms

    In 2025, he coauthors "Unified Breakdown Analysis for Byzantine Robust Gossip" with Renaud Gaucher and Aymeric Dieuleveut, published in the proceedings of the International Conference on Machine Learning. The paper shows an upper bound on the number of adversaries that decentralized algorithms can tolerate and introduces CS+, a robust aggregation rule with a near-optimal breakdown.

  7. June 2026

    Soil fauna image analysis

    In June 2026, he is first author of the bioRxiv preprint "Linking automated image analysis to ecological inference: high-throughput monitoring of soil fauna", written with eight coauthors. The work combines classical computer vision with fine-tuned foundation models to detect invertebrates in soil scans, and it releases an open dataset of about 600 soil scans and more than 8,000 labeled invertebrate occurrences across nine taxa.

  8. As of October 2026

    Researcher at Inria Grenoble

    As of October 2026, he is a researcher (chargé de recherche) in the Thoth team at Inria Grenoble. He focuses on projects at the interface of machine learning and ecology, and is co-PI, with Vincent Miele, of the XLBiod MIAI chair on explainable machine learning methods for ecology.

Career

  1. NowInriaResearcher (chargé de recherche)
  2. 2016–2022EPFLResearch Assistant, Postdoc

Education

  1. 2021InriaPhD
  2. 2018EPFLMaster's degree in computer science
  3. 2016École Polytechnique

From the biography.

  1. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other universityIncludes Hadrien Hendrikx. The Massachusetts Institute of Technology trained or employed 44 of the 187 AI major-prize winners, more than any other university. We compare doctorate and 5+ year employment links, career timing and current workplaces for 187 major-prize and 1,111 best-paper winners.
  2. People shaping AI work at a median of 3 organizations over a career: 3 for university starters and institute and lab starters, versus 2 for company startersIncludes Inria. Across 6,853 people shaping AI with a dated work role, we estimate a median of 3 parent organizations over a career (mean 3.6), comparing role groups, starting sectors and starting decades from the 1950s to 2020–2026.
  3. The company share of AI researchers’ top papers was 31% in 2020–2024 and 8% in 1970–1974Includes Inria. We estimate where the people shaping AI worked when their most-cited papers appeared: the company share was 31% in 2020–2024 and 8% in 1970–1974, and Google led all organizations in 2020–2024.
  4. Google has the highest combined h-index among AI organizationsIncludes Inria. We estimate that Google holds the highest combined h-index at 8,016, while universities account for 73% of the deduplicated total across AI organizations.

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 Hadrien Hendrikx yet. Send us a free-licensed photo.

  1. Hadrien Hendrikx di.ens.fr. Checked 21 October 2026.
  2. Dynamic Safe Interruptibility for Decentralized Multi-Agent Reinforcement Learning arxiv.org. Checked 21 October 2026.
  3. Statistically Preconditioned Accelerated Gradient Method for Distributed Optimization proceedings.mlr.press. Checked 21 October 2026.
  4. NeurIPS 2021 Awards neurips.cc. Checked 21 October 2026.
  5. Announcing the NeurIPS 2021 Award Recipients blog.neurips.cc. Checked 21 October 2026.
  6. A Continuized View on Nesterov Acceleration for Stochastic Gradient Descent and Randomized Gossip arxiv.org. Checked 21 October 2026.
  7. Beyond spectral gap: the role of the topology in decentralized learning proceedings.neurips.cc. Checked 21 October 2026.
  8. Unified Breakdown Analysis for Byzantine Robust Gossip proceedings.mlr.press. Checked 21 October 2026.
  9. Linking automated image analysis to ecological inference: high-throughput monitoring of soil fauna | Sciety sciety.org. Checked 21 October 2026.