Mathieu Blondel
Research scientist at Google DeepMind; former lead developer of JAXopt and former scikit-learn core developer
- Research scientist, Google DeepMind1
Mathieu Blondel has been a research scientist at Google DeepMind since 2019, and from 2013 to 2019 was a research scientist at NTT's Communication Science Laboratories. Blondel was the lead developer of JAXopt, a library of hardware-accelerated, batchable and differentiable optimizers in JAX, and is first author of the paper on efficient and modular implicit differentiation that the JAXopt README cites for its implicit differentiation framework; Blondel also contributed to Optax and now contributes to gemma, a JAX library for using and fine-tuning Gemma models. With Vincent Roulet, Blondel wrote The Elements of Differentiable Programming, a book reviewing the fundamental concepts of differentiable programming, and Blondel is a co-author of the 2011 scikit-learn paper and an emeritus scikit-learn maintainer.1234578
Work
JAXopt (external site: github.com)Open-source code
Was the lead developer of JAXopt, a library of hardware-accelerated, batchable, differentiable optimizers in JAX, and is first author of "Efficient and Modular Implicit Differentiation" (2021), which the JAXopt README cites for its implicit differentiation framework.2349
gemma (JAX library for Gemma) (external site: github.com)Open-source code
Listed on Blondel's software page as a current contributor to gemma, an Apache-2.0 JAX library for using and fine-tuning Gemma models.26
The Elements of Differentiable Programming (external site: arxiv.org)Published research
Co-author with Vincent Roulet of the 2024 book reviewing differentiable programming from optimization and probability perspectives.5
scikit-learn (external site: github.com)Open-source code
Former core developer, co-author of the 2011 JMLR scikit-learn paper, and listed among the emeritus maintainers on the project's About page.27810
Earlier roles
- Research scientist (2013–2019), NTT Communication Science Laboratories1
This profile covers public professional work only, from the sources below. It is not an endorsement or a ranking. To correct or remove it, use the correction form.