BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Penn Engineering Events - ECPv6.15.18//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Penn Engineering Events
X-ORIGINAL-URL:https://seasevents.nmsdev7.com
X-WR-CALDESC:Events for Penn Engineering Events
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20200308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20201101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20210314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20211107T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20220313T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20221106T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20211012T153000
DTEND;TZID=America/New_York:20211012T163000
DTSTAMP:20260406T125601
CREATED:20211005T161145Z
LAST-MODIFIED:20211005T161145Z
UID:5548-1634052600-1634056200@seasevents.nmsdev7.com
SUMMARY:CIS Seminar:"David V.S. Goliath: the Art of Leaderboarding in the Era of Extreme-Scale Neural Models"
DESCRIPTION:Scale appears to be the winning recipe in today’s leaderboards. And yet\, extreme-scale neural models are still brittle to make errors that are often nonsensical and even counterintuitive. In this talk\, I will argue for the importance of knowledge\, especially commonsense knowledge\, and demonstrate how smaller models developed in academia can still have an edge over larger industry-scale models\, if powered with knowledge. \nFirst\, I will introduce “symbolic knowledge distillation”\, a new framework to distill larger neural language models into smaller commonsense models\, which leads to a machine-authored KB that wins\, for the first time\, over a human-authored KB in all criteria: scale\, accuracy\, and diversity. Next\, I will introduce a new conceptual framework for language-based commonsense moral reasoning\, and discuss how we can teach neural language models about complex social norms and human values\, so that the machine can reason that “helping a friend” is generally a good thing to do\, but “helping a friend spread fake news” is not. Finally\, I will discuss an approach to multimodal script knowledge\, which leads to new SOTA performances on a dozen leaderboards that require grounded\, temporal\, and causal commonsense reasoning.
URL:https://seasevents.nmsdev7.com/event/cis-seminardavid-v-s-goliath-the-art-of-leaderboarding-in-the-era-of-extreme-scale-neural-models/
ORGANIZER;CN="Computer and Information Science":MAILTO:cherylh@cis.upenn.edu
END:VEVENT
END:VCALENDAR