BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Towards a complete theory of representation learning and generaliz
 ation in linear Bayesian neural networks
DTSTART:20240126T131500
DTEND:20240126T141500
DTSTAMP:20261002T004116Z
UID:92645e5eab5c14699855fe710cd6634056dda4472b7f85111b9e8718
CATEGORIES:Conferences - Seminars
DESCRIPTION:Jacob Zavatone-Veth (Harvard) \nUnderstanding how representat
 ion learning affects generalization is among the foremost goals of modern 
 deep learning theory. In this talk\, I will discuss the significant recent
  progress that has been made towards understanding perhaps the simplest to
 y model for deep representation learning: deep linear Bayesian neural netw
 orks. For these models\, we can obtain a precise asymptotic characterizati
 on of generalization and representation learning\, and in some cases even 
 obtain closed-form solutions at finite size. I will conclude by commenting
  on remaining gaps in our understanding\, and on transferrability of insig
 hts to nonlinear models. 
LOCATION:GA 3 21 https://plan.epfl.ch/?room==GA%203%2021
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR