BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Efficient Structured Prediction on Long Texts - Mrinmaya Sachan (E
 TH Zurich)
DTSTART:20221028T110000Z
DTEND:20221028T120000Z
UID:TALK189119@talks.cam.ac.uk
CONTACT:Michael Schlichtkrull
DESCRIPTION:Abstract: \n\nVast majority of past NLP research has focused o
 n domains such as tweets\, blogs\, Wikipedia and news articles. However\, 
 documents in several other domains of interest such as scientific articles
 \, legal proceedings or novels and textbooks are substantially longer. Lon
 g documents pose a significant computational challenge to typical NLP mode
 ls. In this talk\, I will focus on structured prediction over long texts\,
  particularly motivated by the problem of scaling coreference resolution m
 odels to long documents. State of the art end-to-end coreference models us
 e expensive span representations and antecedent prediction mechanisms. The
 se approaches are expensive both in terms of their memory requirements as 
 well as compute time\, and are particularly ill-suited for long documents.
  I would describe a succession of recent efforts from our group in scaling
  these models using a) efficient structured span selection which relies on
  the intuition that most spans of interest in typical span selection tasks
  are syntactic constituents\, b) token level span representations and near
 est neighbor sparsification for more efficient antecedent prediction\, and
  c) autoregressive structured prediction which models structures as a sequ
 ence of actions in a dynamic action space using large language models. Thi
 s is joint work with Tianyu Liu\, Yuchen (Eleanor) Jiang\, Raghuveer Thiru
 kovalluru\, Kumar Shridhar\, Nicholas Monath and Ryan Cotterell.\n\nBio: \
 n\nMrinmaya Sachan is an Assistant Professor of Computer Science at ETH Zu
 rich. His research is in the area of Natural language processing and the i
 nterface of Machine learning and Education. Prior to this position\, Mrinm
 aya was a Research Assistant Professor at TTI Chicago. Before that\, he re
 ceived a Ph.D. from the Machine Learning Department at CMU and a B.Tech. i
 n Computer Science from IIT Kanpur where he received an Academic Excellenc
 e Award. He has received several awards for his work\, including an outsta
 nding paper award at ACL 2015\, an IBM PhD fellowship\, the Siebel scholar
 ship and the CMU CMLH fellowship. His current research is funded by grants
  from the Swiss National Science Foundation\, the ETH Zurich foundation an
 d Haslerstiftung.\n\nTopic: NLIP Seminar\nTime: Oct 28\, 2022 12:00 PM Lon
 don\n\nJoin Zoom Meeting\nhttps://cl-cam-ac-uk.zoom.us/j/91073515866?pwd=U
 nJmTER6dmZLeWpPOUo0VUJBOGxYQT09\n\nMeeting ID: 910 7351 5866\nPasscode: 64
 6960
LOCATION:Virtual (Zoom)
END:VEVENT
END:VCALENDAR