BEGIN:VCALENDAR
VERSION:2.0
CALSCALE:GREGORIAN
PRODID:UW-Madison-Physics-Events
BEGIN:VEVENT
SEQUENCE:1
UID:UW-Physics-Event-6812
DTSTART:20220112T170000Z
DTEND:20220112T180000Z
DTSTAMP:20260414T154349Z
LAST-MODIFIED:20220110T152344Z
LOCATION:Online: https://uwmadison.zoom.us/j/92870556335
SUMMARY:A gentle introduction to new ideas in modern ML\, Physics ∩ 
 ML Seminar\, Kangwook Lee\, UW Madison ECE and CS
DESCRIPTION:In this talk\, I will cover three new (and orthogonal) ide
 as in modern ML -- transformers\, diffusion models\, and differentiabl
 e simulation.  Transformers are modern neural network architectures\, 
 which enabled recent breakthroughs in various AI fields.  Diffusion mo
 dels and differentiable simulations are for data generation and invers
 e problems.  I will focus on sharing high-level ideas without many tec
 hnicalities\, assuming minimal ML background.  I will conclude the tal
 k with potential applications of the introduced ideas for physics rese
 arch.
URL:https://www.physics.wisc.edu/events/?id=6812
END:VEVENT
END:VCALENDAR