Archive for Paris

πάντες γὰρ οἱ λαβόντες μάχαιραν ἐν μαχαίρᾳ ἀπολοῦνται [verbatim]

Posted in Uncategorized with tags , , , , , , , , , , , , on February 23, 2026 by xi'an

Last week, Quentin Duranque, a young neo-fascist militant (and incidentally a data science student) was killed during a street fight between far-right and far-left groups, in conjunction with far-right protests against a conference at Science Po’ Lyon by the LFI euro-deputee Rima Hassan. This tragic event reminded me of the mirrored death of the far-left militant and Science Po’ student Clément Méric in June 2013, who also died during a street fight between far-right and far-left groups in Paris. Beyond the ghastly nonsense of such deaths, alas made possible by the will to engage into meaningless physical clashes (hence the quote from Matthew as a title), the political exploitation of the killings by French parties was appalling if predictable. The almost identical statements from leaders of both far-right (FN) and far-left (LFI) parties following both deaths are striking in that respect:

“Le climat créé par l’extrême droite conduit à ce genre de drame (…) L’extrême droite porte une responsabilité morale dans ce qui s’est passé. C’est le résultat de la banalisation des idées d’extrême droite (…) Les fascistes ont tué.” Jean-Luc Mélenchon, Juin 2013

” Ce drame est le résultat direct du climat de haine entretenu contre les patriotes (…) Ce meurtre est le résultat d’années de diabolisation de nos militants. Ceux qui désignent nos militants comme des ennemis prennent une responsabilité terrible.” Jordan Bardella, Fév. 2026

“Je condamne évidemment cet acte odieux (…) Il est scandaleux de tenter de récupérer politiquement ce drame (…) Ce drame est le résultat de l’affrontement de groupuscules violents. Le Front national n’a rien à voir avec ces individus.” Marine Le Pen, Juin 2013

“La violence ne peut jamais être une solution. Toute violence est condamnable, quelles qu’en soient les victimes ou les auteurs (…) La responsabilité est individuelle, pas collective.” Jean-Luc Mélenchon, Fév. 2026

“Il est irresponsable d’instrumentaliser ce drame. La responsabilité est celle des individus impliqués.” Manuel Bompart, Fév. 2026

mostly Monte Carlo [20/02]

Posted in Statistics with tags , , , , , , , , , , , , , , , , on February 16, 2026 by xi'an

A new episode of our mostly Monte Carlo seminar, very soon coming near you (if in Paris):

On Friday 20/02/26, from 3-5pm at PariSanté Campus

15h: Paul Mangold (École Polytechnique, Palaiseau)

Convergence and Linear Speed-Up in Stochastic Federated Learning

In federated learning, multiple users collaboratively train a machine learning model without sharing local data. To reduce communication, users perform multiple local stochastic gradient steps that are then aggregated by a central server. However, due to data heterogeneity, local training introduces bias. In this talk, I will present a novel interpretation of the Federated Averaging algorithm, establishing its convergence to a stationary distribution. By analyzing this distribution, we show that the bias consists of two components: one due to heterogeneity and another due to gradient stochasticity. I will then extend this analysis to the Scaffold algorithm, demonstrating that it effectively mitigates heterogeneity bias but not stochasticity bias. Finally, we show that both algorithms achieve linear speed-up in the number of agents, a key property in federated stochastic optimization.

16h: Alain Durmus (École Polytechnique, Palaiseau)

A Mixture-based Framework for Guiding Diffusion Models

Inverse problems—such as image restoration from noisy or incomplete measurements and musical source separation—are ill-posed, making Bayesian approaches with learned generative priors especially appealing. Diffusion models provide powerful priors, but existing posterior sampling methods often rely on crude likelihood-gradient approximations and heavy task-specific tuning. In this talk, I will introduce a novel principled approach specifically designed to overcome these limitations. The core contribution of this approach is the construction of a mixture approximation of intermediate posterior distributions defined by the diffusion model. The sampling is carried out sequentially via Gibbs sampling, a Markov Chain Monte Carlo method, using a careful data augmentation scheme. Gibbs sampling is employed here due to its simplicity and theoretical guarantees, allowing for exact conditional updates at each iteration, thus ensuring stability and efficiency. One key advantage of the presented algorithm is its flexibility: it adapts to varying levels of computational resources by adjusting the number of Gibbs iterations. Consequently, substantial performance gains can be achieved by increasing inference-time computational effort. I will present extensive experimental results demonstrating empirical performance across diverse image restoration tasks, involving both pixel-space and latent-space diffusion models, and showcase its successful application in musical source separation.

 

Paul Deheuvels (1948-2026)

Posted in Books, Mountains, University life with tags , , , , , , , , , , , , , , , , , , , , , , , on February 5, 2026 by xi'an

I just learned that Paul Deheuvels died last week. He was head of the statistics department at Université Pierre & Marie Curie when I was a lecturer there (1987-1992). And a professor of statistics and probability there for his entire career, from 1974 to 2013. While we were working on completely different topics, and had divergent political views, we kept an amicable connection throughout the years, as we were living close enough to meet once in a while at the local farmers’ market and have a quick chat. When I was considering my options for pursuing a PhD, while writing a Master thesis on commutative Lie algebras with his father, René Deheuvels, he was one of the people who offered to supervise me (on a topic related with empirical processes, which were his forte). Later, he suggested I wrote a very short introduction to Bayesian statistics in the Que sais-je? collection (where he published three books) and, when my manuscript was found too theoretical,  he proposed to publish an expanded version in the Economica collection, which ended up being the French version of The Bayesian Choice. I also remember a very kind email he sent me while I was recovering from my climbing accident, in 2013, as a fellow climber. (For a long while, he would drive to Fontainebleau once or twice a week for bouldering.) As alluded (to) above, Deheuvels’ political views were very conservative, with a proximity to the far-right Club de l’Horloge and a mix of vieille France traditionalism and libertarianism positions. For sure, he never was shy from embarking in controversies, from defending Allègre’s climato-scepticism to backing Séralini’s  controversial study on the cancerogenic effects of Mosanto’s transgenic maize.

congrats, Doctor Luciano!

Posted in Books, Kids, Statistics, University life with tags , , , , , , , , , , on January 23, 2026 by xi'an

Cédric, le retour ?

Posted in Books, pictures, University life with tags , , , , , , , , , , , , , , on January 12, 2026 by xi'an

Within a non-Euclidean distance of a few issues, Le Monde and Libération both ran an interview of Cédric Villani, not truly a coïncidence since he recently published Leçons de mathématique joyeuse, a general public book on mathematics (I did not read), derived from a series of conferences he gave in Paris ten years ago. With a chapter on the best and the worst errors of (Henri) Poincaré, where he states that statistics is a late-comer in the mathematics pantheon because “it is complicated, it is even completely counter-intuitive.” However, there is very little on mathematics and research in these interviews, as the interviewers are clearly more interested in the politics of Villani, a former deputé and unsuccessful candidate for the City of Paris mayorship. Although he is now teaching in both Lyon and Rennes, since he currently lives in a nearby village that happens to be very close to my family house in Morbihan.