Rémi Flamary

POT Python Optimal Transport is 10 years old this week 🎂. Did not think when I pushed the first commit 10 years ago that it would still be here or have this impact (~11M dowloads on PyPI). Open Source Software is a marathon not a sprint. Thanks to all contributors and users (who cite it) 🥳.

I'm reading Transmetropolitan for the fifth time and it is still one of the best comic I ever read. Also it is crazy how it touches on everything happening today.

Short thread asking @nytimes.com and other MSM to stop interviewing the same small handful of people every time something math-related comes up. Mathematicians are dismayed when AI scoops them on the results? How about look for advice from someone familiar with getting scooped and outpowered.

Approximating convex functions is hard; input-convex neural networks (ReLu networks with positive weights) are convex, but for finite depth, they are not universal approximators. www.gpeyre.com/blog/2026/10...

Trois mathématiciennes lauréates du prix Jeunes talents France Fondation L'Oréal-UNESCO 2026 🏆 Félicitations à Brune Massoulié, Claudia Rella et Sonia Mazelet ! Découvrez leur parcours ▶️ bit.ly/4if8666

In a general-public piece, I discuss how ideas can go from a conversation with an AI tool to a mathematical proof. It leads me to wonder, how much plagiarism is there in the mathematical genius of bi AI?

Impressive use of Gromov-Wasserstein and Wasserstein Procruste to align between embeddings without supervision.

So sad to hear that the iconic, trailblazing, Apollo 11 moon landing software engineering lead, Margaret Hamilton, is no longer with us. Love this image of her, standing next to a print out of the software that landed Buzz Aldrin and Neil Amstrong on the moon! 👩‍🔬🔭🛰️🧪 ℹ️: www.bbc.co.uk/news/article...

Our second call for NeurIPS Paris 2026 volunteers will open tomorrow TUE Oct 6th, 2pm CET. We are still seeking for PhD students and postdocs to assist during the conference in exchange for complimentary registration. Application form: forms.gle/bBFUyUNngtCc...

Cette BD est incroyable et les perso très attachants, pour le prix de 2 trajets de métro, on a une bd de @bouletcorp.bsky.social (qui sera finie si on en achète assez...) et on soutient un site original et une super idée pour rémunérer des créateurs correctement.

Use #bibCheck if you are a French academic: it's an online tool that audits bibliographies and reports retracted or inexistent references, developed in collaboration with @cnrs-inist.bsky.social (see slide 25 ut3-toulouseinp.hal.science/hal-05646645...). Otherwise check github.com/gianlucasb/h...

Nihar Shah did a heroic experiment for TMLR: he spent 20-25 hours over two weeks interviewing authors of seemingly low-quality submissions about their own papers. He confirmed what we all suspected: people submitting these papers have *no idea* what is going on in them.

New year, new semester, new course, new textbook: davidpicard.github.io/mldl/mldl-bo... 👀 Still in draft form, but not in a bad shape.

NeurIPS is approaching and registration for the Paris site will be opened on Monday Sept 14th, Anywhere on Earth. Registration happens at neurips.cc, each of the 3 sites has its own registration process. We'll be happy to see you in Paris!

Today, we will send a preference form to all NeurIPS authors. Please discuss your preferences with your co-authors and submit the form by September 30, 2026. If your paper is accepted, your presenter and location preferences will help us allocate papers fairly across our three conference locations.

It seems like there's a new round of "stochastic parrots has been proven wrong" discourse going on, so I'm going to re-up this: medium.com/@emilymenonb... >>

The first call for NeurIPS Paris 2026 volunteers is open! We are seeking for PhD students and postdocs to assit during the conference in exchange for complimentary registration. Application form: forms.gle/HGDUzS7y7yvq... 1st call deadline: September 10th

I will be using this Figure at the begining of my courses and insist that I'm back to paper exam this year. It shows the scores of AI using student for homework increased but decreased for exam. papers.ssrn.com/sol3/papers....

www.youtube.com/watch?v=saoF...

We are happy to announce that 28 workshops have been accepted for the Paris event, as part of the 102 accepted NeurIPS workshops: blog.neurips.cc/2026/08/10/a... They will take place on Sat Dec 12 + Sun Dec 13, 2026 (for Paris) The suggested deadline for Workshop submissions is close (Aug 29th)!

This is a magnificent collection of visual pedagogy. Will definitely steal (and cite) some ideas for my slides. Strongly recommended especially to the ones that don't follow @gabrielpeyre.bsky.social on social media.

I'm burnt out of discussing "AI". I don't care for the press releases from big companies. I hate having to comment immediately. I dislike "hot takes". I'm worried we, as scientists, are losing touch with scientific values. We've become customers, salivating at ads marketed to investors. 1/15

I'm happy to announce the #NeurIPS2026 workshop TS-LIMITS (Generalization for Time Series in Tight Settings: Latency, Inference, Memory, prIvacy and susTainability), which will be hosted in Paris, France 🇫🇷, Dec 12–13 2026! 🔗 Check out details and call for papers: ts-limits.github.io

Back to paper exam for my machine learning course next year (thanks LLMs for ruining lab reports BTW). But at least I'm having fun designing ML "skill" questions. My neighbors can probably hear "wicked witch of the west" cackling this morning.

Excited to announce that our workshop "Bridging Optimal Transport, Learning and Structured Data: Toward Geometric Distributional Learning" has been accepted at #NeurIPS2026 in Paris. ⏰Deadline: Aug 29, AoE 🔗Website with more informations: gddl-neurips-2026.github.io

Jean-Edouard avait un sens des priorités plus développé que celui du bas peuple Ma première BD, les fables bucoliques autogérées, est sortie ! 🦊 A retrouver dans votre librairie indé préférée (lien de commande en bio ou sur www.placedeslibraires.fr/livre/978241...)

A new (?) concept: the interactiv logo: www.gpeyre.com/ot4ml/

Une rupture d'égalité devant la loi qui n'a pas sa place dans un état de droit...

We just released POT 0.9.7 with many new OT solvers including lazy and sparse exact solvers, HD Gaussians, sliced OT plans, Spectral-Grassmann OT, batch exact OT with Proximal point, BSP-OT, unbalanced 1D and semidiscrete. This is a big one! Thanks to all contributors! github.com/PythonOT/POT...

I’ll be at ICML presenting our paper! Come say hi and check out our poster on Thursday 9th (Hall A, poster #3710). Huge thanks to @simonroschmann.bsky.social @paulkrz.bsky.social, @qbouniot.bsky.social and @zeynepakata.bsky.social for this great collaboration !

“I have done all of the things that one is supposed to do to earn a tenure-track position. And I have done approximately 85% of them by typing prompts into a large language model and then moderately editing the output.” 🤣🤣🤣 brilliant 👌🏽 open.substack.com/pub/inprepar...

Meteo france has an over-fitting problem (or a bug)

Yes. But I fear the train has left the station, there is nothing we can do. Not using LLMs is too big of a disadvantage. 1/2

This article in Le Monde about French research in general and the CNRS in particular has a few depressing graphs www.lemonde.fr/sciences/art...

Can compute alone replace strong and original research ideas? No luck so far on my side

🚀 Excited to announce the release of skada v0.6.0! This new release brings important additions to the library and continued work to keep the ecosystem compatible with Sklearn!

Congratulations to @tgnassou.bsky.social for the SSFAM PhD prize. Théo has done both very good (I might be biased) ML research but also contributed a lot to the open source community with SKADA, skada-bench and contributions to POT. Plus for potential recruiters it's a pleasure to work with him

I'm thrilled to have been awarded the thesis prize by the SSFAM 🥳 Thank you to the jury for the recognition you've given me 🙏 A big thanks to @rflamary.bsky.social and Alex to supporting me in this PhD! I'm looking forward to present my work at CAp 2026 in Montpellier, see you there :)

It would not have been possible withou POT, big thanks to the team pythonot.github.io

@gabrielpeyre.bsky.social has written a book that will (again) have a huge impact on the field of OT for ML. Strongly recommended!

The alpha version of my new book "Optimal Transport for Machine Learners" is out, with in particular an online version with interactive figures www.gpeyre.com/ot4ml/

Yep I'm been playing with the BSP-OT solver and it is quite fast. The following interpolation is done on 300k samples and only takes a few seconds (but can only find permutations). Generating the GIF/animation is slower than solving the problem.

BSP-OT is now available in POT main branch! github.com/PythonOT/POT (👏 @baptiste-genest.bsky.social )

NeurIPS 2026 Call for Workshop Proposals is still open! For the first time, NeurIPS includes a European venue — Paris 🇫🇷, co-organised by several ELLIS Members. 📅 Deadline: June 6 👉 https://neurips.cc/Conferences/2026/CallForWorkshops @neuripseurope.bsky.social

So honored and humbled to have won this award! Thanks David ☺️

Open call for a 1+ year PostDoc Fellowship at Institut Polytechnique de Paris financed by the Hi! Paris AI center. You can choose until June 5th your own research project and just need a support letter of a mentor at IP Paris. hi-paris.fr/wp-content/u...