Greg Durrett

@gregd_nlp

Assistant Professor at UT Austin. I do NLP most of the time. he/him

Vrijeme pridruživanja: prosinac 2017.

Tweetovi

Blokirali ste korisnika/cu @gregd_nlp

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @gregd_nlp

  1. Prikvačeni tweet
    24. lip 2019.

    Over the past two years at UT I've developed some NLP course materials. I talked to a couple folks at NAACL who had found these useful, so I wanted to give a quick self plug. Still a work in progress, of course: Grad: Ugrad:

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    20. sij

    Excited to share this Youtube playlist of videos we made and where I explain/discuss basics in operating systems! Its loosely based on Three Easy Pieces OS textbook. Playlist includes demos and code walkthroughs!

    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    18. pro 2019.

    Thrilled to share with you the launch of the 2020 Program application! If you know ugrads from trad. marginalized backgrounds interested in summer astro research, please RT this and let them know about TAURUS! Our application link: 1/n

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    17. pro 2019.

    Congrats to our undergrad researcher , who was selected for Honorable Mention for the Computing Research Association's (CRA) Outstanding Undergraduate Researcher Award 2020!

    Poništi
  5. proslijedio/la je Tweet
    9. pro 2019.

    I'm happy to share my comments on the climate for men from my talk:

    Poništi
  6. 9. pro 2019.

    post-weekend ACL paper status

    Poništi
  7. proslijedio/la je Tweet
    4. pro 2019.

    We just published a new dataset for coreference resolution in English literature (as a new layer in LitBank) -- Paper: Data:

    Poništi
  8. proslijedio/la je Tweet
    26. stu 2019.

    Hey academic twitter, when was the last time you updated your webpage photo?

    Poništi
  9. proslijedio/la je Tweet
    21. stu 2019.
    Poništi
  10. proslijedio/la je Tweet
    21. stu 2019.

    For those struggling to come up with a title for their ACL submission... =)

    Poništi
  11. proslijedio/la je Tweet
    14. stu 2019.

    We welcome Dr. Eunsol Choi into our Department as Assistant Professor of Natural Language Processing. Dr. Choi was awarded the Research Fellowship in 2018 and she's working as a visiting research scientist at . 👩👉

    Poništi
  12. 9. stu 2019.

    ...on HotpotQA as well. We also do some human evaluation of whether humans can answer the underlying questions using the chains extracted by our model, and find that the evidence extracted is usually sufficient for them to do so. Full paper here: 4/4

    Prikaži ovu nit
    Poništi
  13. 9. stu 2019.

    We show that these chains are equivalent in quality to annotated supporting facts, but are ordered, generalize to other multi-hop datasets, and don't require manual supervision. Our model achieved state-of-the-art on WikiHop (but has since lost it) and gets strong results ... 3/n

    Prikaži ovu nit
    Poništi
  14. 9. stu 2019.

    A sequential pointer network extracts the sentences. Unlike other approaches that have been used on datasets like HotpotQA, we don't use annotated supporting facts. Instead, we derive reasoning chains based on entity relations to use at train time. 2/n

    Prikaži ovu nit
    Poništi
  15. 9. stu 2019.

    New(ish) preprint available: , work by Jifan Chen and Shih-Ting Lin. We tackle multi-hop QA as a problem of extracting a discrete reasoning chain over sentences, then doing QA (with BERT) over those sentences. (1/n)

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    8. stu 2019.

    does not have a standard benchmark for interpretability. I am stoked to announce ERASER: the first-ever effort on unifying and standardizing NLP tasks with the goal of interpretability.

    Prikaži ovu nit
    Poništi
  17. proslijedio/la je Tweet
    4. stu 2019.

    UNT is hiring in / ML / AI (2 openings). Please reach out if interested (DM or email), or we can chat at

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet
    3. stu 2019.

    Great turn out for Workshop at @emnlp2019. The room is fully loaded and the poster sessions are super well attended! Shoutout to speakers Dan Roth & co-organizers Shahram !

    Poništi
  19. proslijedio/la je Tweet
    31. lis 2019.

    Check out our paper "Evaluating Lottery Tickets Under Distributional Shifts" (), to appear as an oral at the workshop at . Rather surprisingly, we show lottery tickets generalize to unseen domains despite high distributional shifts. (1/3)

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    2. stu 2019.

    Our paper “Learning Dense Representations for Entity Retrieval” will be presented on Monday. I contributed to this effort two summers ago during my internship with & Daniel Gillick and am proud how it turned out. 1/2

    Prikaži ovu nit
    Poništi
  21. 1. stu 2019.

    (P.S. Aditya will be applying for PhD programs this cycle, so you should hire him!)

    Prikaži ovu nit
    Poništi

Čini se da učitavanje traje već neko vrijeme.

Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

    Možda bi vam se svidjelo i ovo:

    ·