BEGIN:VCALENDAR VERSION:2.0 PRODID:-//events.la.psu.edu//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York BEGIN:STANDARD DTSTART:20201101T020000 TZOFFSETFROM:-0400 TZOFFSETTO:-0500 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD BEGIN:DAYLIGHT DTSTART:20200308T020000 TZOFFSETFROM:-0500 TZOFFSETTO:-0400 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT END:VTIMEZONE BEGIN:VEVENT UID:19316-57439651ac440efaaa08410d3a8b479d@events.la.psu.edu DTSTAMP:20260923T203853Z DTSTART;TZID=America/New_York:20261002T090000 DTEND;TZID=America/New_York:20261002T103000 SUMMARY:Center for Language Science Speaker Series: Jonathan Dunn, Ph.D. DESCRIPTION:\nLanguage Models enable important new methodologies for stu dying linguistic variation\, especially in domains like syntax. The prob lem is that these models are now embedded within speech communities and provide an increasing amount of our linguistic experience. This talk foc uses on some emerging sociolinguistic consequences of LLMs: their measur able skew across geographic dialects and the possibility for population- specific models\, homogenization caused by machine-assisted writing\, mo del perception of variants resulting in synthetic language attitudes\, a nd the lack of adequate meta-data as an ultimate cause of model collapse . These are critical issues in language technology that need to be resol ved before they cause lasting damage.\n\n \;\n\nFor more details: ht tps://events.la.psu.edu/event/cls-speaker-series-dunn/ X-ALT-DESC;FMTTYPE=text/html:

Language Models enable important new methodologies for studying linguistic variation, e specially in domains like syntax. The problem is that these models are n ow embedded within speech communities and provide an increasing amount o f our linguistic experience. This talk focuses on some emerging sociolin guistic consequences of LLMs: their measurable skew across geographic di alects and the possibility for population-specific models, homogenizatio n caused by machine-assisted writing, model perception of variants resul ting in synthetic language attitudes, and the lack of adequate meta-data as an ultimate cause of model collapse. These are critical issues in la nguage technology that need to be resolved before they cause lasting dam age.

For more details: https://events.la.psu.edu/event/cls -speaker-series-dunn/

URL:https://cls.la.psu.edu/news-events/cls-speaker-series/ LOCATION:Foster Auditorium, 102 Paterno Library END:VEVENT END:VCALENDAR