BEGIN:VCALENDAR VERSION:2.0 PRODID:ILLC Website X-WR-TIMEZONE:Europe/Amsterdam BEGIN:VTIMEZONE TZID:Europe/Amsterdam X-LIC-LOCATION:Europe/Amsterdam BEGIN:DAYLIGHT TZOFFSETFROM:+0100 TZOFFSETTO:+0200 TZNAME:CEST DTSTART:19700329T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU END:DAYLIGHT BEGIN:STANDARD TZOFFSETFROM:+0200 TZOFFSETTO:+0100 TZNAME:CET DTSTART:19701025T030000 RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT UID:/NewsandEvents/Archives/2018/newsitem/9912/15- May-2018-Computational-Linguistics-Seminar-Arianna -Bisazza DTSTAMP:20180423T133600 SUMMARY:Computational Linguistics Seminar, Arianna Bisazza ATTENDEE;ROLE=Speaker:Arianna Bisazza (Leiden Univ ersity) DTSTART;TZID=Europe/Amsterdam:20180515T110000 LOCATION:Room F1.15, Science Park 107, Amsterdam DESCRIPTION:What makes recurrent neural networks w ork so well for next word prediction? Do neural tr anslation models learn to extract linguistic featu res from raw data and exploit them in any explicab le way? In this talk I will give an overview of re cent work, including my own, that aims at answerin g these questions. I will also present recent expe riments on the importance of recurrency for captur ing hierarchical structure with sequential models. X-ALT-DESC;FMTTYPE=text/html:\n
What makes rec urrent neural networks work so well for next word prediction? Do neural translation models learn to extract linguistic features from raw data and expl oit them in any explicable way? In this talk I wil l give an overview of recent work, including my ow n, that aims at answering these questions. I will also present recent experiments on the importance of recurrency for capturing hierarchical structure with sequential models.
URL:http://projects.illc.uva.nl/LaCo/CLS/ END:VEVENT END:VCALENDAR