Zhu JQ

@JQ_Zhu

I study psychology with some good friends. What fun! Postdoc at Warwick.

Shanghai, China
Vrijeme pridruživanja: studeni 2010.

Tweetovi

Blokirali ste korisnika/cu @JQ_Zhu

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

  1. proslijedio/la je Tweet
    9. ruj 2019.

    I believe I have written more papers than Alan Turing + John Nash! Numbers of papers alone is a wrong misleading metric. Please focus instead on writing good papers that advance the field, help the world, and that you’ll be proud of when you look back in 20 or 50 years.

    Poništi
  2. proslijedio/la je Tweet

    Google Scholar seems to be altering scholarly citation patterns. Citations are getting more concentrated: the same few papers get cited over and over, has found. People lazily cite whatever papers the search engine ranks highly.

    Poništi
  3. proslijedio/la je Tweet

    Tremendous, stunning paper by , finding that doctors are influenced by recent, salient adverse events, in a way that is not consistent with Bayesian updating and that harms patients. The availability heuristic in action.

    Poništi
  4. 25. srp 2019.

    presenting my poster today at , we found evidence for amortised inference & sequential sampling until a threshold in post-decisional bias on perception:

    Poništi
  5. proslijedio/la je Tweet
    22. lip 2019.

    Just got refused a Canadian visa I needed to present my poster at . I believe that not everybody fully realizes how much visa issues hurt diversity in academia. So here goes my thread on the topic.

    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    19. lip 2019.

    Oldies but goldies: C. Stein, A bound for the error in the normal approximation to the distribution of a sum of dependent random variables, 1972. Uses Stein’s lemma to derive rates in the central limit thm.

    Poništi
  7. proslijedio/la je Tweet
    8. lip 2019.

    Humans also overfit to the training environment

    Tweet je nedostupan.
    Prikaži ovu nit
    Poništi
  8. proslijedio/la je Tweet
    20. svi 2019.

    A Galton board can physically create Normal distributions just from beads falling randomly. What about other distributions? A multiplicative version exists and can produce log-normal distributions and it's amazing.

    Prikaži ovu nit
    Poništi
  9. proslijedio/la je Tweet
    26. tra 2019.

    Replay of experience in the hippocampus is not necessarily a strict reflection of behavior, but can be random! Work from the Csicsvari lab

    Poništi
  10. proslijedio/la je Tweet
    23. tra 2019.

    "Particle filtering: the first 25 years and beyond" by Simon Godsill. A 5 pages historical account

    Poništi
  11. proslijedio/la je Tweet

    Yoshua Bengio, Geoffrey Hinton and Yann LeCun, the fathers of , receive the 2018 for conceptual and engineering breakthroughs that have made deep neural networks a critical component of computing today.

    Poništi
  12. proslijedio/la je Tweet
    10. ožu 2019.

    Did you know, Brussels-Barcelona is the same distance as Beijing-Shanghai? But the train is ~2x slower and ~5x more expensive in Europe where we've chosen planes over long distance trains. For the climate's sake (and our health) we need an ambitious vision for European rail...

    Poništi
  13. proslijedio/la je Tweet
    10. ožu 2019.

    on this day, exactly 10 years ago, this astonishing paper was published. The pseudo-marginal approach is truly remarkable and allows exact Bayes when using an approximate likelihood. Sounds incredible? My intro blog post: The paper:

    Poništi
  14. proslijedio/la je Tweet
    27. velj 2019.

    I'm looking to hire a full-time research assistant: Unfortunately only US citizens can apply for this position.

    Poništi
  15. proslijedio/la je Tweet

    Statistics doesn't identify a unique truth, it just identifies what truths are _consistent_ with the data you have observed. Empirical evidence out of context is not a sufficient condition for truth.

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet

    This is how Gibbs samplers actually work and this gif is perfect.

    Tweet je nedostupan.
    Poništi
  17. proslijedio/la je Tweet
    19. velj 2019.

    4 postdoc positions to investigate how the mind uses sampling to approximate Bayesian inference — at University of Warwick with Adam Sanborn. Great positions, very interesting topic, with a great person. Check it out.

    Poništi
  18. proslijedio/la je Tweet
    14. velj 2019.

    Our new paper is out! In conditions of state uncertainty, the model-based system retrospectively infers the past, hence guiding model-free credit assignment. With Mehdi Keramati, Peter Dayan and Ray Dolan.

    Poništi
  19. 9. velj 2019.
    Poništi
  20. proslijedio/la je Tweet
    4. velj 2019.

    We're open-sourcing a multiagent environment based on the highly popular card game Hanabi, & an agent based on the Dopamine framework! w/ & al. Here's a short post about the project:

    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:

    ·