Anna Rogers

@annargrs

Post-doc at UMass. Worrying about meaning representations, NLP methodology, and computational social science.

Vrijeme pridruživanja: rujan 2018.

Tweetovi

Blokirali ste korisnika/cu @annargrs

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

  1. Prikvačeni tweet
    19. srp 2019.

    Here's a summary post on problems with huge models that dominate these days. I put together several different discussion threads with/by , , , , , .

    Poništi
  2. prije 10 sati

    Join Yoshua Bengio in this petition to allow remote conference presentations: "...so that anyone be free to decide... whether the benefits of attending on site outweigh the negative consequences of the trip - both for climate and for family life."

    Poništi
  3. 4. velj

    Had a fantastic time at : a rare case of an invited talk that turns into a productive discussion in the end. and James Pustejovsky have great students, I wish I could steal a few! Thanks for the invitation.

    Poništi
  4. 30. sij

    Ok, to be clear: this is not about a paper under review, it already got published. But it's definitely not the only one, so it's good to have this discussion.

    Prikaži ovu nit
    Poništi
  5. 30. sij

    A paper mentions in the abstract that one of its contributions is supplementing an existing dataset with translations into more languages. On page 3 turns out it was just Google-translated. Does that count as an actual resource contribution?

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    9. sij

    The Workshop on Linguistic and Neurocognitive Resources (LiNCR) addresses how linguistic resources, datasets and tasks link to human language cognition. Submissions: 8pp, 4pp, abstracts Deadline: 2020-02-17

    Poništi
  7. 18. sij

    When I see 6 assigned reviews for , I think that I'll handle it by:

    Poništi
  8. 16. sij

    Some of our findings are relevant in the context of the head ablation studies (, ), the easy-dataset problem (, , , , , and many others), and the Great Attention Debate (, , ).

    Prikaži ovu nit
    Poništi
  9. 16. sij

    A belated blog post for our BERTology EMNLP paper (by Olga Kovaleva, Alexey Romanov, yours truly and ). My favorite experiment in this work is showing that for most GLUE tasks BERT works pretty well even *without pre-training*!

    Prikaži ovu nit
    Poništi
  10. 15. sij

    The CFP for is out! People who will be resubmitting ICML submissions owe a beer to , who pointed out the date problem.

    Poništi
  11. 14. sij

    This should be titled "All the words that you heard DL people throw around at conferences, but were too embarrassed to ask". A great list, with links to papers included.

    Poništi
  12. 14. sij

    Discussion between the chairs is ongoing, stay tuned :)

    Poništi
  13. 10. sij

    Excited to announce that I'll serve as publicity chair for (together with Xu Ruifeng on Weibo). We're taking over from and - pretty big shoes to fill!

    Poništi
  14. 10. sij

    Open Reading Benchmark is out! If you're tired of GLUE, here's a fresh look at learning a wide range of reasoning strategies within one task (extractive reading comprehension).

    Poništi
  15. 7. sij

    Nearly forgot: today is the last Christmas of the season! The Orthodox one. If you haven't heard the Carol of the Bells in its original language (Ukrainian), here it is. The song has nothing to do with bells, but it proves that some Slavs do have vowels.

    Poništi
  16. 6. sij

    An opportunity highly recommended. Great group, great city, and also a great coffee machine :)

    Poništi
  17. 3. sij

    Something is wrong with my news feed today. - Koalas are burning alive as you are reading this - New Samsung phones are coming, and are obviously the best use of your money Dear Aussies, any donate links recommendations? The koalas seem to be low on PR staff.

    Poništi
  18. 31. pro 2019.

    The quote is from Andrew Hodge's book on Enigma. The Imitation Game movie is based on it, but, being a movie... you know... it couldn't really get into all the maths, engineering, and philosophy. With a deep appreciation of where we are now, happy New Year!

    Prikaži ovu nit
    Poništi
  19. 31. pro 2019.

    I think Turing says that, at least theoretically, we *can* get rid of any task we get bored of (as it must be repetitive => potentially definable as a machine task). If we can't define it that way, we don't really understand it => not boring.

    Prikaži ovu nit
    Poništi
  20. 31. pro 2019.

    Dear coders, Alan Turing says we've got the best jobs: > This process of constructing instruction tables should be very fascinating. There need be no real danger of it ever becoming a drudge, for any processes that are quite mechanical may be turned over to the machine itself.

    Prikaži ovu nit
    Poništi
  21. proslijedio/la je Tweet
    23. pro 2019.

    Love my German collaborator's OOO: "I am currently on vacation. As research shows that working during holidays can be detrimental for health and well-being (de Bloom, Geurts & Kompier, 2012), I will not check my emails to mentally detach from work (Sonnentag & Fritz, 2015)."

    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:

    ·