Chaitanya (Chay) Ryali

@wrong_whp

PhD student in Computer Science at UC San Diego. Interested in computational cognitive science, neuroscience, cognitively/neurally inspired ML.

Vrijeme pridruživanja: lipanj 2018.

Tweetovi

Blokirali ste korisnika/cu @wrong_whp

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

  1. proslijedio/la je Tweet
    16. sij

    New preprint: Locality sensitive hashing with synaptic weights learned in a biologically-plausible way, SOTA results on unsupervised similarity search based on hashing for MNIST and CIFAR-10, and discussion of the computational role of olfactory networks.

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

    Meta-reviewer boldly claims that our entire theorem is wrong by providing a "counterxample" wherein they claim S=[1 0; 0 -2] to be a positive definite matrix, while denouncing our claims to be false!! 🤯Are you sure you want to be meta-reviewing? Please be more considerate.

    Poništi
  3. proslijedio/la je Tweet
    7. sij

    (4) Differences in the degree to which men-led and women-led teams undercite women are only partly explained by the structure of authors’ social networks; after accounting for social networks, women-led teams cite relatively equitably, while men-led teams still overcite men.

    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    7. sij

    (2) The undercitation of women-led papers is driven largely by the citation behavior of men-led teams; (3) Undercitation of women-led papers is increasing over time, particularly within the reference lists of men-led teams;

    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    7. sij

    New from Extent & drivers of gender imbalance in neuroscience reference lists . w/ , , Teich, Linn. Covering issues of bias, ethics, & social behavior vital to all fields grappling with their own visions of an equitable future.

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    13. pro 2019.
    Prikaži ovu nit
    Poništi
  7. proslijedio/la je Tweet
    13. pro 2019.

    Yu: humans overestimate environment volatility, use suboptimal decision policies, and underestimate prior mean rewards. (paper: ) .

    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    13. pro 2019.

    Angela Yu now presenting on the important consideration that humans have idiosyncratic learning and decision making - they assume the environment dynamics are changeable, even if they are truly are fixed (paper: ).

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    13. pro 2019.

    Angela Yu is giving a great talk on "Features or Bugs: Synergistic Idiosyncrasies in Human Learning and Decision-Making." Join us in the Ballroom C or catch the rest of the talk via livestream!

    Poništi
  10. proslijedio/la je Tweet
    10. pro 2019.

    Come to to hear about BioHash - a bio-inspired algorithm motivated by fruit fly olfactory system. SOTA results on unsupervised similarity search based on hashing for MNIST and CIFAR-10. Joint work of and .

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

    I will be presenting my internship project with cortex tonight 6:30 on deep Bayesian bandit for personalized online recommendation, where we attempt to mitigate dataset bias issue with exploration! Poster #345, stop by and discuss!

    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    6. pro 2019.

    [ Workshop Paper Preview] Inspired by FlyHash, a fruit-fly-inspired hashing algorithm, and Dmitry Krotov present BioHash, that can make use of data to learn its hash codes. Papers / Schedule at:

    Poništi
  13. proslijedio/la je Tweet
    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    21. kol 2019.

    Researchers currently don't have a forum for sharing updated thoughts on their past papers. introduces ML Retrospectives, a new NeurIPS workshop that wants to change that.

    Poništi
  15. proslijedio/la je Tweet
    Odgovor korisnicima i sljedećem broju korisnika:
    Poništi
  16. proslijedio/la je Tweet
    1. kol 2019.

    Why do ML researchers keep applying information theory to deterministic settings? It doesn’t make any sense.

    Poništi
  17. proslijedio/la je Tweet
    16. lip 2019.
    Odgovor korisnicima

    Reckon that last point is the key one, tbh. Don’t see yourself as a gatekeeper, but rather a colleague trying to help someone produce better work.

    Poništi
  18. proslijedio/la je Tweet
    24. svi 2019.

    On the choice of deep learning hyperparameters

    Prikaži ovu nit
    Poništi
  19. proslijedio/la je Tweet
    14. ožu 2019.

    Those were the days, my good friend We wrote gradients down by hand They weren't too complex Our losses were convex You’re too young, you won't understand

    Poništi
  20. proslijedio/la je Tweet
    1. ožu 2019.

    Exciting news. If a few more big universities show the same spirit, this whole scam will fall apart. It's only a matter of time.

    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:

    ·