Colloquia & Seminars

Title: Abelian duality via derived geometry

Quantum Field Theory and Physical Mathematics Seminar Speaker: Owen Gwilliam, UMass Amherst Title: Abelian duality via derived geometry Abstract: We discuss how to synthesize differential cohomology and the BV formalism to describe generalized Maxwell theories (or abelian p-form gauge theories), and how this framework allows a succinct formulation of abelian duality. Given time, we will discuss how these methods apply to the 6d self-dual 2-form gauge theory that appears as part of the 6d N=(1,0) and (2,0) superconformal theories known as abelian tensor multiplets. This is joint work in progress with Chris Elliott, Ingmar Saberi, and Brian Williams.  

Title: Geometry and Quantum Theory Seminar

Joint Math/CMSA Geometry and Quantum Theory Seminar Speaker: Vasily Krylov, Harvard

Title: Separation of timescales controls feature learning and overfitting in large neural networks

New Technologies in Mathematics Seminar Speaker: Pierfrancesco Urbani, Universite Paris-Saclay, CNRS, CEA, Institut de physique theorique Title: Separation of timescales controls feature learning and overfitting in large neural networks Abstract: To understand the inductive bias and generalization capabilities of large, overparameterized machine learning models, it is essential to analyze the dynamics of their training algorithms. Using dynamical mean field theory we investigate the learning dynamics of large two-layer neural networks. Our findings reveal that, for networks with a large width, the training process exhibits a separation of timescales phenomenon. This leads to several key observations: 1. The emergence of a slow timescale linked to the growth in Gaussian/Rademacher complexity of the network; 2. An inductive bias favoring low...

Title: CMSA Q&A Seminar: Hugh Woodin, Harvard

CMSA Q&A Seminar Speaker: Hugh Woodin, Harvard Title: Truth, proof, and AI