3 papers
stat.ML2026
A Fourier perspective on the learning dynamics of neural networks: from sample complexities to mechanistic insights
Fabiola Ricci, Claudia Merger, Sebastian Goldt
Neural networks trained with gradient-based methods exhibit a strong simplicity bias: they learn simpler statistical features of their data before moving to more complex features.…
stat.ML2025
Feature learning from non-Gaussian inputs: the case of Independent Component Analysis in high dimensions
Fabiola Ricci, Lorenzo Bardone, Sebastian Goldt
Deep neural networks learn structured features from complex, non-Gaussian inputs, but the mechanisms behind this process remain poorly understood. Our work is motivated by the obse…
math.DG2024
Isoperimetric profile function comparisons with Integral Ricci curvature bounds
Jihye Lee, Fabio Ricci
We prove comparison results for the Isoperimetric profile function in the setting of manifolds with integral bounds on the Ricci curvature. We extend previous work of Ni and Wang a…