BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//https://www.dfp.ubc.ca//NONSGML iCalcreator 2.41.92//
CALSCALE:GREGORIAN
METHOD:PUBLISH
UID:33613632-6462-4333-b663-353261633335
X-WR-RELCALID:efc09d74-9c93-479e-a94f-485231ddccde
X-WR-TIMEZONE:America/Vancouver
X-WR-CALNAME:Learning to Unlearn: Robust and Efficient Machine Unlearning f
 or Large Foundation Models - Moontae Lee\, Assistant Professor\, Universit
 y of Illinois Chicago
BEGIN:VTIMEZONE
TZID:America/Vancouver
TZUNTIL:20270314T100000Z
BEGIN:STANDARD
TZNAME:PST
DTSTART:20241103T020000
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
RDATE:20251102T020000
RDATE:20261101T020000
END:STANDARD
BEGIN:DAYLIGHT
TZNAME:PDT
DTSTART:20250309T020000
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
RDATE:20260308T020000
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:d3dc974a-1edc-497d-84ea-3be08d74aaa0
DTSTAMP:20260109T045800Z
CLASS:PUBLIC
CREATED:20250704T225445Z
DESCRIPTION:Abstract: As large foundation models gain massive popularity\, 
 the ability to remove sensitive information without fully retraining from 
 scratch becomes crucial. This talk introduces three recent works for machi
 ne unlearning that can erase data or knowledge selectively from pretrained
  models. Instance-wise unlearning leverages adversarial examples with weig
 ht importance metrics to erase unwanted information without access to the 
 pretraining dataset. Then we examine gradient ascent for post-training unl
 earning in large language models\, demonstrating how targeted token sequen
 ces can be forgotten…
DTSTART;TZID=America/Vancouver:20250715T121500
DTEND;TZID=America/Vancouver:20250715T131500
LAST-MODIFIED:20250704T225931Z
LOCATION:UBC Vancouver Campus\, Fried Kaiser (KAIS) building\, Room 2020/20
 30\, 2332 Main Mall
SUMMARY:Learning to Unlearn: Robust and Efficient Machine Unlearning for La
 rge Foundation Models - Moontae Lee\, Assistant Professor\, University of 
 Illinois Chicago
TRANSP:OPAQUE
URL:https://www.dfp.ubc.ca/index.php/event/learning-unlearn-robust-and-effi
 cient-machine-unlearning-large-foundation-models-moontae-lee
END:VEVENT
END:VCALENDAR