On 17 July 2025 at 5 p.m., Prof. Dr. Guido Montúfar, Professor of Mathematics and Statistics & Data Science at the University of California (USA), will speak about “Deep Learning theory: What we know, what we are learning, and what remains unclear”. No registration is necessary.
Deep learning has revolutionized artificial intelligence and a wide range of applied domains, driving transformative progress in computer vision, language processing, and scientific discovery. This talk surveys the vibrant and rapidly evolving landscape of deep learning theory—an effort to uncover the mathematical foundations of learning with neural networks. We will review key theoretical insights into optimization dynamics, implicit biases of learning algorithms, and the generalization behavior of deep models—highlighting connections to classical learning theory, high dimensional statistics, and approximation theory. Along the way, we will discuss some of the major successes in analyzing overparameterized regimes, as well as open challenges in understanding feature learning and generalization under moderate overparameterization. The talk will also spotlight emerging phenomena such as benign overfitting, grokking, and delayed generalization, illustrating the depth and complexity of ongoing research questions that challenge traditional notions.
About the speaker: Prof. Dr. Guido Montúfar
Guido Montúfar is a Professor of Mathematics and of Statistics & Data Science at the University of California, Los Angeles, and leads the Mathematical Machine Learning research group at the Max Planck Institute for Mathematics in the Sciences, MPI MiS. He earned dual Diplom degrees in Mathematics and Physics from TU Berlin in 2007 and 2009, and completed his Dr. rer. nat. in Mathematics at Leipzig University and MPI MiS in 2012. Prior to joining UCLA in 2017, he held postdoctoral positions in the Department of Mathematics at Penn State and at MPI MiS. Professor Montúfar’s work bridges applied mathematics, statistics, data science, and machine learning, and he has led numerous initiatives at the intersection of these fields. He serves as a core Principal Investigator in the SECAI Zuse School of Excellence in AI (Leipzig–Dresden) and leads the project “Combinatorial and Implicit Approaches to Deep Learning” within the DFG Priority Program on the Theoretical Foundations of Deep Learning (FoDL). His work has been recognized with honors including the ERC Starting Grant (2017), the NSF CAREER Award (2021), and the Alfred P. Sloan Research Fellowship (2022).