sada-lab.ch

Next in our AI for Science and Science(s) of AI talk series: Friedemann Zenke (FMI) on how internal prediction shapes learning (29 October), followed by Siddhartha Mishra (ETH Zurich) on neural operators and foundation models for physics (3 November).

Mark your calendars!

Dandelion: A Spherical Flower for Neural Simulation of Planetary Dynamics and On Observation Time for Recovering Latent Hawkes Networks have been accepted to NeurIPS 2026. Congratulations to Till and Giovanni, and to Jonas, Michele and Lorenzo! 

We were happy to welcome Richard for a secondary school internship at SADA. Special thanks to Geoffrey and Jose from SciCore for giving him a great tour of the data center!

Congratulations to Till and Alex — FLOWERs: A Warp Drive for Neural PDE Solvers was accepted to ICML 2026 and selected as a Spotlight paper! 

Gaussian Surrogates for Poisson Imaging: Some Theoretical and Empirical Results was accepted to ICIP 2026, well done Alex and Lorenzo!

Our paper 🍦 Icecream: High-Fidelity Equivariant Cryo-Electron Tomography has been accepted for publication in Acta Crystallographica Section D, check out the preprint here! 

Prof. Dokmanić has been awarded an ERC Consolidator Grant for the project PhaseShift​ – Phase-Space Foundations for Scientific Machine Learning, congratulations Ivan!

Opening announcements coming very soon!

Prof. Dokmanić has been selected for a Swiss National Science Foundation project grant for Event-based representation and learning of stochastic systems on graphs, congratulations Ivan! 

Cheng has successfully defended his PhD and will join Stéphane Mallat’s team at École Normale Supérieure in Paris as a postdoctoral researcher. Congratulations, Cheng. We wish you all the best!

Cheng's paper Spring-Block Theory of Feature Learning in Deep Neural Networks was featured in reasearch highlight articles on phys.org, myscience.ch and by the University of Basel.  Congratulations!