At the same time, I cannot write this message without telling that I am **very** worried about the current administrative evolution of our profession which is evolving quickly towards a system absolutely not calibrated to our duties with a purely accounting PoV.

(*) directeur de recherche is roughly equivalent to full prof w/o mandatory teaching duty in France. This is an incredible tenured civil servant position, and I am deeply grateful for it. DM me if you have questions about the process to enter CNRS.

Happy to share that I have been promoted to senior research scientist (directeur de recherche (*)) at @cnrs effective today! Grateful to my amazing students, collaborators and mentors who made this journey possible. I will continue my research at the math lab of UniCA (LJAD).

I'm BEYOND THRILLED to tell you that our first original song is now available on all platforms! Plz share if you like it 🙏🏼🙏🏼😊 Cheers

🚀 NeurIPS@Paris is back for a 5th edition at the SCAI. Meet us in central Paris to discuss recent advances in AI! 📆 25th & 26th Nov. 2025 🌐 https://t.co/tgZJ1h2hIF 🎓 Committee: Chloé-Agathe Azencott, @BachFrancis, Claire Boyer, @gerardbiau , @VianneyPerchet , @jeanphi_vert

Described by Philip Gage https://t.co/CQcfVjERt6 as a compression mechanism in 1994 in The C Users Journal

Byte Pair Encoding is a tokenization method that starts with all characters as initial tokens. It iteratively merges the most frequent adjacent byte pairs in the text, adding new tokens to the vocabulary until reaching a predefined size. The output is a sequence of subword tokens

If you are in the Bay Area, consider attending our workshop, "Graph Learning Meets TCS," at the Simons Institute ( https://t.co/vDzzRyR4kA).

q: can we find somewhere the missed attempts to solve pb6 from either google or openai?

SIAM Review on this algorithm (and variants) by Bauschke & Borwein (1996)

Cyclic projection onto convex sets is a method to find a point in the intersection of convex sets. It iteratively projects onto each set in a cycle, and it converges to a point in the intersection under restrictive assumptions.

Introduced by Vaswani et al. "Attention is all you need" https://t.co/xdNmX4lAoU Early occurence for NLP in Bahdanau et al. https://t.co/N928Oo47v0 *Incredible* explanation by @3blue1brown https://t.co/u1gjBicyE7 Annotated (Pytorch) verson of the paper

A single-head unmasked attention layer processes a sequence by assigning relevance scores between elements. It computes these scores to "decide" how much focus each part of the sequence should receive and combines the information.

Congratulations to Dr. Sophie Jaffard @SophieJaffard for a brillant PhD defense! We very lucky with Patricia to have you as a PhD student, and I am sure that you will have a brillant career ahead.

Paper by Hammersley (1956)

Zeros of random polynomials with i.i.d. coefficients exhibit interesting convergence properties. In particular, normal coefficient leads to weak convergence of the empirical measure to the unit circle.

Paper:

The Neural Probabilistic Language Model (Bengio et al., 2000) introduced joint use of word embeddings and neural networks for next-token prediction. It replaced n-gram models by mapping words to a continuous vector space, and the use a standard neural networks architectures.

Neural ODE : https://t.co/MMPaz3VlYX ResNet : https://t.co/3VqJj6XHX5 Do Residual Neural Networks discretize Neural Ordinary Differential Equations? :

ResNet and Neural ODEs are closely related: ResNet uses discrete residual/skip connections, while Neural ODEs generalize this to continuous transformations using ODEs. Neural ODEs *can* be seen as the limit of ResNet as the number of layers approaches infinity.