AlfonsoValencia

@Alfons_Valencia

Bioinformatics, Comp biol, cancer genomes, text mining, strucutral bioinfo. but also: Science publishing, science policy, genome & ethics, science & art

Vrijeme pridruživanja: veljača 2011.

Tweetovi

Blokirali ste korisnika/cu @Alfons_Valencia

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

  1. proslijedio/la je Tweet
    prije 20 sati

    FINAL WEEK of the open call for abstracts of the conference in Paris (26-28 August 2020) "When models, methods & experiments meet the clinic":

    Poništi
  2. proslijedio/la je Tweet

    The Chan Zuckerberg Initiative is announcing 30 grants to patient groups — with the goal of building a model for tackling any rare disease. My story:

    Poništi
  3. proslijedio/la je Tweet
    prije 15 sati

    We need an ambitious : Investments at EU level can deliver the science & innovations that are needed to tackle the challenges too complex to be solved by one country alone, such as the climate emergency, global health (e.g. recent Corona virus outbreak), and security.

    , , i još njih 2
    Poništi
  4. proslijedio/la je Tweet
    prije 8 sati

    My trip to visit calico pointed out this interesting paper reducing the genetic estimate of longevity (below the paltry 15%) due to partner effects; they say assortative partner choice but I think it might well be spouse influenced health influence.

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    prije 14 sati

    The resources we would be lost without - the ELIXIR Core Data Resources paper is now out in the journal of .

    Poništi
  6. proslijedio/la je Tweet
    prije 12 sati

    Today, the ELIXIR EU-funded project kicks off! 📈 Ahead, we have 36 months of stirring change and to move beyond national strategies to an integrated network Click below for more 👇

    Poništi
  7. proslijedio/la je Tweet

    Thomas Wolf on transfer learning at NLPL. How to make embeddings interpretable: visualization strategies

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet

    Transfer learning: pretraining and representation by Thomas Wolf at NLPL. Tutorial URLs. tinyurl/nlplwinter1 , sequential transfer learning

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet

    Anti-vaxxers are essentially a cult. Their latest conspiracy theories are about coronavirus, but they've been at this for years. Read about their strategies here:

    Poništi
  10. "Un sistema de inteligencia artificial fue el primero en alertar del coronavirus de Wuhan" via

    Poništi
  11. proslijedio/la je Tweet
    22. pro 2019.
    Poništi
  12. proslijedio/la je Tweet
    1. velj

    Today's honoree for is Dr. Edward Bouchet. Dr. Bouchet was the first African American to earn a Ph.D. in physics from any American University. He has established a legacy of excellence for all African American physicists.

    Poništi
  13. proslijedio/la je Tweet
    2. velj

    DeepMoji example : beyond word embeddings. Emoji useful for predictions of age, demographics,.. using Twitter. Adversarial training to try to unlearn something, for example gender from age, 3 objectives

    Poništi
  14. proslijedio/la je Tweet
    2. velj
    Poništi
  15. proslijedio/la je Tweet
    2. velj
    Odgovor korisniku/ci

    Some of the underlying assumptions regarding the social group divisions in nameprism are highly problematic. There is implicit cultural stereotyping in the training data. Something like this needs to be addressed.

    Poništi
  16. proslijedio/la je Tweet
    2. velj
    Odgovor korisniku/ci

    Not anymore! This was done before the language model take off

    Poništi
  17. proslijedio/la je Tweet
    1. velj

    🚬Fantàstic article de la a , en aquest cas sobre els beneficis de deixar de fumar. Llegiu-lo!

    Poništi
  18. proslijedio/la je Tweet
    Odgovor korisniku/ci

    Not better, just different. The bogus information is still spreading credibly. They aren't reading or understanding these critical comments. This isn't to say preprints are bad, but this event should point to a weakness that should be addressed.

    Poništi
  19. proslijedio/la je Tweet
    2. velj
    Odgovor korisnicima

    1000s of people walked away from this utterly convinced by its connection to conspiracy theories, saw none of the comments or scitwitter tweets, and will not the retraction. If they did see them, it was "complicated science on both sides". It's a system failure, not a win.

    Poništi
  20. proslijedio/la je Tweet
    1. velj

    Based on the content of my mentions, I feel like I need to further debunk crazy / HIV conspiracy preprint. This is a thread doing so. 1/9

    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:

    ·