BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UM//UM*Events//EN
CALSCALE:GREGORIAN
BEGIN:VTIMEZONE
TZID:America/Detroit
TZURL:http://tzurl.org/zoneinfo/America/Detroit
X-LIC-LOCATION:America/Detroit
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20070311T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20071104T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20261009T150233
DTSTART;TZID=America/Detroit:20261106T150000
DTEND;TZID=America/Detroit:20261106T160000
SUMMARY:Lecture / Discussion:AIM Seminar:  The Geometry of Diffusion Generation: A Log-Density Ridge Perspective
DESCRIPTION:Abstract:  Diffusion models generate data by gradually transforming noise into structured samples. When these samples differ from the training examples\, what determines their geometric organization? In this talk\, I will present a framework based on log-density ridges of the smoothed empirical distribution. These evolving structures provide a geometric reference for analyzing generation relative to a finite training set. I will describe a reach–align–slide mechanism in reverse diffusion: samples enter a ridge neighborhood\, approach the ridge in normal directions\, and move along it relative to nearby training data. Our analysis relates alignment and local tangential motion to directional components of the learned denoising error\, connecting training error to generation geometry. This perspective helps explain how generated samples can organize around structures induced by the data while differing from individual training examples. Synthetic examples and MNIST latent-space experiments illustrate the predicted behavior. The talk will introduce the necessary background on diffusion models and log-density ridges.\n\nThis talk is based on joint work with Yitong Qiu and Molei Tao\, recently accepted at NeurIPS 2026.\n\nContact:  Zhiyan Ding
UID:148881-21905043@events.umich.edu
URL:https://events.umich.edu/event/148881
CLASS:PUBLIC
STATUS:CONFIRMED
CATEGORIES:Applied Mathematics,Mathematics
LOCATION:East Hall - 1084
CONTACT:
END:VEVENT
END:VCALENDAR