BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//wp-events-plugin.com//7.4.13//EN
TZID:America/New_York
X-WR-TIMEZONE:America/New_York
BEGIN:VEVENT
UID:867@fds.yale.edu
DTSTART;TZID=America/New_York:20250626T130000
DTEND;TZID=America/New_York:20250626T140000
DTSTAMP:20250916T142153Z
URL:https://fds.yale.edu/events/fds-seminar-akhil-premkumar-uchicago-an-in
 formation-theoretic-view-of-machine-learning/
SUMMARY:FDS Seminar: Akhil Premkumar (UChicago)\, "An information theoretic
  view of machine learning"
DESCRIPTION:\nAbstract: Diffusion models serve as a bridge between generat
 ive AI and information theory. These models have demonstrated a remarkable
  ability to learn high-dimensional continuous distributions\, like images 
 and video\, from relatively small training datasets. They can do this beca
 use they learn the ensemble statistics of the entire dataset\, allowing th
 em to identify long wavelength correlations between the given samples. In 
 other words\, these models are extremely good at compression. Drawing on
  ideas from thermodynamics and information theory\, we introduce a method 
 to measure the information content of a neural network\, called neural ent
 ropy. Measurements of neural entropy in diffusion models reveal that the a
 mount of new information the model learns from each additional sample dimi
 nishes with the total number of training samples. We will also draw some c
 onnections between probabilistic modeling and games of chance.\n\n\n\nBase
 d on: https://arxiv.org/abs/2409.03817v2 and ongoing work.\n\n\n\nSpeake
 r Bio: I am a postdoc at the Kavli Institute for Cosmological Physics at 
 the University of Chicago. I did my PhD in theoretical physics at Universi
 ty of California San Diego\, where I studied stochastic processes that eme
 rge from combining quantum mechanics and gravity.\n\n\n\nHosted by John So
 us.\n
CATEGORIES:FDS Events,Seminar Series,Summer Seminar
LOCATION:Yale Institute for Foundations of Data Science & Webcast\, 219 Pro
 spect Street\, 13th Floor\, New Haven\, CT\, 06511\, United States
X-APPLE-STRUCTURED-LOCATION;VALUE=URI;X-ADDRESS=219 Prospect Street\, 13th 
 Floor\, New Haven\, CT\, 06511\, United States;X-APPLE-RADIUS=100;X-TITLE=
 Yale Institute for Foundations of Data Science & Webcast:geo:0,0
END:VEVENT
BEGIN:VTIMEZONE
TZID:America/New_York
X-LIC-LOCATION:America/New_York
BEGIN:DAYLIGHT
DTSTART:20250309T030000
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
END:DAYLIGHT
END:VTIMEZONE
END:VCALENDAR