Au département d’informatique de l’ENS, le CNRS et Inria main dans la main dans l’effort. #LEtapeduTour Bravo Olivier Cappé!

Antonio Orvieto and I are planning to co-advise a student from this program. If you are interested in optimization for deep learning, please apply!

Bravo Julien!!

The blog is back in 2024, with a first post on unraveling spectral properties of kernel matrices! francisbach.com/spectrum-kerne…

🙎‍♂️‍🔬 #30minScience | Aujourd’hui, au centre Inria de #Paris, Umut U. Simsekli (Sierra Team) nous présente une demi-heure de science consacrée aux #Fractales : ”Fractals in Machine Learning”. 🔎 Découvrez en plus sur Umut et ses recherches dans : di.ens.fr/umut.simsekli/…

Bravo Adrien!

How fast is gradient descent, *for real*? Some (partial) answers in this new blog post on scaling laws for optimization. francisbach.com/scaling-laws-o…

#ERCStG 🏆 | Félicitations à Adrien Taylor, chargé de recherche @inria au centre Centre Inria de Paris, membre de l'équipe-projet commune Sierra Team (CNRS 🌍 CNRS Sciences informatiques École normale supérieure | PSL), lauréat d'une ERC Starting Grant 👏 Découvrez son portrait et son projet 👉 buff.ly/3YlTBU6

🎓#PhDs | Félicitations à Juliette Murris et Alice Rogier de l’équipe HeKA research team (Université Paris Cité, @CRCordeliers, Inserm) ainsi qu'à Oumayma Bounou et Céline Moucer des équipes WILLOW et Sierra Team (CNRS Sciences informatiques, École normale supérieure | PSL) qui ont soutenu leur #thèse en octobre 🎉👏 !

Bravo Bertille et Marc!

My book is (at last) out, just in time for Christmas! A blog post to celebrate and present it: francisbach.com/my-book-is-out/

🔴 Inria est présent à la conférence "AI,Science and society" à Institut Polytechnique de Paris, prélude au #SommetActionIA et sous la direction du Pr. Michael Jordan (UC Berkeley-Inria) ! L'événement a été inauguré ce matin par Michael Jordan et cet après-midi par Clara Chappaz 🗣

[🔴#IAScienceSociety] Francis Bach (Centre Inria de Paris-@ENS_ParisSaclay) participe au symposium "The Mathematics of #MachineLearning" aux côtés de Gabriel Peyré, Gilles Louppe (Université de Liège), Gersende Fort (CNRS 🌍) et Vianney Perchet (ISAE Institut Polytechnique de Paris).

Michael Jordan is indeed one of the greatest thinkers in the history of AI 🐐 Economics, incentives (mechanism design), information flow, creativity, cooperation, greed and power struggles are important topics that we crucially need to understand better for the benefit of

Characterizing finely the decay of eigenvalues of kernel matrices: many people need it, but explicit references are hard to find. This blog post reviews amazing asymptotic results from Harold Widom (1963!) and proposes new non-asymptotic bounds. francisbach.com/spectrum-kerne…

A free book: Learning Theory from First Principles by Francis Bach It covers a bunch of key topics from machine learning (ML) theory and practice, such as: - Math basics - Supervised learning - Generalization, overfitting & adaptivity - Tools to design learning algorithms -

For good probability predictions, you should use post-hoc calibration. With Eugène Berta, Michael Jordan, and Francis Bach we argue that early stopping and tuning should account for this! Using the loss after post-hoc calibration often avoids premature stopping. 🧵1/

I’ll be presenting our paper at COLT in Lyon this Monday at the Predictions and Uncertainty workshop — come say hi if you're around! 👋 Check out David Holzmüller's thread below 👇 #COLT2025

Happy to have our recent papers on conformal prediction with e-values presented at COLT by my advisor Francis Bach! Full details here: 📚arxiv.org/abs/2503.13050 📚arxiv.org/abs/2505.13732 #COLT2025

Tired of lengthy computations to derive scaling laws? This post is made for you: discover the sharpness of the z-transform! francisbach.com/z-transform/