@ncourty.bsky.social
Professor in Computer Science. Love and hate AI Optimal Transport Affinicionado Head of Obelix group @Irisa
I see Sirat in poster... I hope this was your choice (assuming those are photos of a Cinema) !
16.02.2026 14:52 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0We are recruiting four positions connected to Machine Learning, Statistical Learning, and AI for Science in the Applied Mathematics department at รcole polytechnique. Join our vibrant community at IP Paris and Hi! Paris IA center. List below๐งต tinyurl.com/3jpw9t26
06.02.2026 07:56 โ ๐ 11 ๐ 19 ๐ฌ 1 ๐ 0Maybe it is also LLM generated ?
29.01.2026 17:12 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0
"On meurt dans mon Universitรฉ", par la prรฉsidente de l'Universitรฉ Paul Valรฉry ร Montpellier. Parce que les baisses de financement des universitรฉs ce sont aussi des conditions de travail si dรฉgradรฉes qu'elles en deviennent intenables
27.01.2026 14:51 โ ๐ 45 ๐ 42 ๐ฌ 0 ๐ 1beware of the abstract submission deadline !
23.01.2026 13:01 โ ๐ 1 ๐ 0 ๐ฌ 0 ๐ 0North Brittany ?
27.12.2025 22:23 โ ๐ 0 ๐ 0 ๐ฌ 1 ๐ 0๐น Job alert: 3 PhD positions in AI, Earth Observation, and Science-Policy interface
๐ Vannes ๐ซ๐ท & Ispra ๐ฎ๐น
๐
Apply by Jan 15th
๐ https://www-obelix.irisa.fr/job-offers/
One of those internships is on Gromov $\delta$-hyperbolicity for GNNs, and will be cosupervised together with Nicolas, myself and Laetitia Chapel. Take a look and spread the words !
07.11.2025 13:45 โ ๐ 10 ๐ 3 ๐ฌ 1 ๐ 0Halloween aprรจs l'heure
05.11.2025 11:09 โ ๐ 2 ๐ 0 ๐ฌ 0 ๐ 0so true....
23.10.2025 09:34 โ ๐ 0 ๐ 0 ๐ฌ 0 ๐ 0
#Distinction ๐| Charlotte Pelletier, laurรฉate d'une chaire #IUF, dรฉveloppe des mรฉthodes dโintelligence artificielle appliquรฉes aux sรฉries temporelles dโimages satellitaires.
โก๏ธ www.ins2i.cnrs.fr/fr/cnrsinfo/...
๐ค @irisa-lab.bsky.social @cnrs-bretagneloire.bsky.social
I love how they managed to distill the spirit of the first Alien movie. And the Eye creature .... !
23.09.2025 07:53 โ ๐ 1 ๐ 0 ๐ฌ 1 ๐ 0Huge fan too. Brace yourself for episode 5 ^^
18.09.2025 09:49 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0Trying hard to decouple my interest for the scientific questions behind AI and this....๐ฎโ๐จ
07.09.2025 08:59 โ ๐ 10 ๐ 0 ๐ฌ 1 ๐ 1
Abstract: Under the banner of progress, products have been uncritically adopted or even imposed on users โ in past centuries with tobacco and combustion engines, and in the 21st with social media. For these collective blunders, we now regret our involvement or apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not considered a valid position to reject AI technologies in our teaching and research. This is why in June 2025, we co-authored an Open Letter calling on our employers to reverse and rethink their stance on uncritically adopting AI technologies. In this position piece, we expound on why universities must take their role seriously toa) counter the technology industryโs marketing, hype, and harm; and to b) safeguard higher education, critical thinking, expertise, academic freedom, and scientific integrity. We include pointers to relevant work to further inform our colleagues.
Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI (black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAIโs ChatGPT and Appleโs Siri, we cannot verify their implementation and so academics can only make educated guesses (cf. Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al. 2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA).
Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe.
Protecting the Ecosystem of Human Knowledge: Five Principles
Finally! ๐คฉ Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia:
doi.org/10.5281/zeno...
We unpick the tech industryโs marketing, hype, & harm; and we argue for safeguarding higher education, critical
thinking, expertise, academic freedom, & scientific integrity.
1/n
I always appreciate @cwarzel.bsky.social's takes on AI!
๐
Distributional Reduction paper with H. Van Assel, @ncourty.bsky.social, T. Vayer , C. Vincent-Cuaz, and @pfrossard.bsky.social is accepted at TMLR. We show that both dimensionality reduction and clustering can be seen as minimizing an optimal transport loss ๐งต1/5. openreview.net/forum?id=cll...
27.06.2025 07:44 โ ๐ 33 ๐ 9 ๐ฌ 1 ๐ 1A subtle combination of the three ? With a pinch of regrets about what could have been perfected ?
16.05.2025 21:55 โ ๐ 3 ๐ 0 ๐ฌ 0 ๐ 0
We have been reworking the Quickstart guide of POT to show multiple examples of OT with the unified API that facilitates access to OT value/plan/potentials. It allows to select regularization/unbalancedness/lowrank/Gaussian OT with just a few parameters. pythonot.github.io/master/auto_...
26.03.2025 07:39 โ ๐ 32 ๐ 11 ๐ฌ 0 ๐ 0maybe, in the end, you are the innie @chriswolfvision.bsky.social ?
22.03.2025 16:25 โ ๐ 2 ๐ 0 ๐ฌ 1 ๐ 0
Solutions to the PAWL problem in 1D for different amounts of mass to be transported
โ๏ธ One for all and all for one โ๏ธ
Efficient computation of PArtial Wasserstein distances on the Line (PAWL)
is accepted to @iclr-conf.bsky.social
Joint work with Laetitia Chapel: we introduce an ๐(๐ ๐๐๐ ๐) solver for partial Optimal Transport (OT) in 1D
openreview.net/forum?id=kzE...
๐งต 1/2
Slicing Unbalanced Optimal Transport
Clรฉment Bonet, Kimia Nadjahi, Thibault Sejourne, Kilian FATRAS, Nicolas Courty
Action editor: Benjamin Guedj
https://openreview.net/forum?id=AjJTg5M0r8
#transport #outliers #optimal
Nice ! But no all the flow matching methods rely on an independent coupling, some use mini-batch OT also, see e.g. arxiv.org/abs/2304.14772 or arxiv.org/abs/2302.00482
02.12.2024 13:29 โ ๐ 6 ๐ 0 ๐ฌ 1 ๐ 0
