Super excited to be teaching (1/3rd) the Advanced Topics in #DeepLearning course with Stefan Sommer, and two other teachers, with a focus on graph neural networks. Emergency alarm went off in the building after my first slide. That must be a good omen. No? 😅

Thorben Pieper-Sethmacher, Frank van der Meulen, Aad van der Vaart: On a class of exponential changes of measure for stochastic PDEs arxiv.org/abs/2409.08057 arxiv.org/pdf/2409.08057

Oldies but goldies: Sathamangalam Srinivasa Varadhan, On the behavior of the fundamental solution of the heat equation with variable coefficients, 1967. Shows that the heat kernel on a manifold is approximately a geodesic Gaussian kernel for a small time. onlinelibrary.wiley.com/doi/10.1002/cp…

Looking forward to this seminar by trevordarrell Berkeley AI Research next week!

I am recruiting PhD students for 2025!😃 You want to reveal the geometric signatures of natural and artificial intelligence, and unlock the mysteries of the brain?🤖🧠 Apply to the UCSB Geometric Intelligence Lab✨ This is the view you'd have from... well, your desk🌴

Kalman filter defines recursively an estimator of the parameters of a Gaussian dynamical process. The basic method to control navigation systems. en.wikipedia.org/wiki/Kalman_fi…

Shape evolution…. We knew it Stefan Sommer

The president of the National Academy of Sciences, Marcia McNutt, has published welcome an essay promising to remain politically neutral and mitigate the political polarization of science. Some notable quotes from the article: "[S]cientists need to better explain the norms and

Intervista con mio fratello Alex Bronstein riguarda l'episodio obbrobrioso a Cagliari. Il silenzio dell Università degli Studi di Cagliari ufficiale e' pure una vergogna. cagliaritoday.it/attualita/inte…

Check out the ELLIS Pre-NeurIPS Fest event today in... 🇩🇰 Copenhagen! ELLIS Unit Copenhagen ELLIS Unit Copenhagen is holding their event at Pioneer Centre for AI showcasing #NeurIPS posters and other Denmark-affiliated papers in #AI and #ML. More info: bit.ly/4fRFrAh

Gradient descent on particles’ positions (Lagrangian) is equivalent to an advection linear PDE on the density of particles (Eulerian). en.wikipedia.org/wiki/Lagrangia…

Interested in a post-doc position in theoretical computer science? Want to find out why they call it "Wonderful Copenhagen"? Look here! Application deadline January 10. Hiring in multiple areas including geometry, complexity, learning, and privacy. jobportal.ku.dk/videnskabelige…

Collection of insane and fun facts about SQLite. Let's go! SQLite is the most deployed and most used database. There are over one trillion (1000000000000 or a million million) SQLite databases in active use. It is maintained by three people. They don't allow outside

The gradient field defines the steepest descent direction. The gradient flow dynamic defines a segmentation of the space into attraction basins of the local minimizers. en.wikipedia.org/wiki/Gradient en.wikipedia.org/wiki/Gradient_…

I'll be hiring PhDs and postdocs in 2025 (and beyond) for two new research projects, MorphAI and DIGIFABA (at UCPH, funded by Novo Nordisk Foundation ) If you have a solid background in math/stats/cs (logical or, but depth is required), wish to join a growing research group in

The space of compact sets in a metric space is a compact set for the Hausdorff metric. Hausdorff convergence is weak and does not preserve topology, dimension, length, or path-connectedness. It preserves connectedness and convexity. en.wikipedia.org/wiki/Hausdorff…

Open phd-positions in my group, including computer vision for detection and segmentation in plant images candidate.hr-manager.net/ApplicationIni…

Score Matching Riemannian Diffusion Means arxiv.org/abs/2502.13106 Score matching is applicable for learning the reverse dynamics of diffusion processes, for bridges etc., but also for fast computation of means on manifolds. By learning the score of the Riemannian Brownian motion,

New version of our paper on parallel transport on matrix manifolds arxiv.org/abs/2408.06054 , to appear in SIMAX

Neural Guided Diffusion Bridges arxiv.org/abs/2502.11909 w/ Gf Yang Frank van der Meulen We introduce a new bridge simulation method that combines the guided proposals of Frank van der Meulen and Moritz Schauer with an additional correction drift term parametrized by a learnable neural