Max Puelma Touzel

@mptouzel

Husband&dad • IVADO postdoc @ • Fig1&7 *T*heory in physics, biology and learning • coupling silos for collective effects• plays well with neurons

Montréal, Québec
mptouzel.github.io
Vrijeme pridruživanja: travanj 2017.
4 fotografije ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @mptouzel

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

  1. Prikvačeni tweet
    20. pro 2019.

    Did you see "A new hope"? Not StarWars, but our lively panel discussion with Bengio, Fiete, , Lillicrap, , Savin, at (end of TOC; starts at 42:27). All workshop content online!

    9 proslijeđenih tweetova 34 korisnika označavaju da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    16. sij

    Since this comes up a lot, a quick run though of the testable, falsifiable, science that supports a human cause of recent trends in global mean temperature.

    29 replies 626 proslijeđenih tweetova 992 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. 18. sij

    Nice sqrt scaling of # items recalled to # items stored over the requisite 2 orders of magnitude

    Our free recall paper is finally out. Phys. Rev. Lett. 124, 018101 (2020) - Fundamental Law of Memory Recall
    Poništi
  4. proslijedio/la je Tweet
    15. sij

    Some of ’s most fascinating work happens at the intersection of AI and neuroscience. Our paper published in today is a good example. It takes inspiration from our recent work in Reinforcement Learning to shed new light on how dopamine works in the brain.

    More exciting news today: an example of how AI and neuroscience continue to propel each other forward. (1/2)
    Prikaži ovu nit
    5 replies 139 proslijeđenih tweetova 524 korisnika označavaju da im se sviđa
    Poništi
  5. 15. sij

    Recent example of NeuroAI approach for RL.

    Poništi
  6. 25. pro 2019.

    Umpa lumpa brain theory

    We explain a homunculus not by positing a second homunculus within it that performs the same activity, but by positing a team of smaller, less talented, and more specialized homunculi - and specifying the ways they cooperate to produce their joint output. -William Lycan
    Prikaži ovu nit
    1 korisnik označava da mu se sviđa
    Poništi
  7. proslijedio/la je Tweet
    22. pro 2019.

    One of my very favorite space images of all time: snowstorm on a comet (plus a whole lotta cosmic ray hits to the sensor). May all your solstices include reflections on the beauty and depth of the exquisite universe around us!

    3.228 proslijeđenih tweetova 11.512 korisnika označava da im se sviđa
    Poništi
  8. 22. pro 2019.

    "[Physicists] too have been brought up inside a discipline, and are as prone as any of us to blunder when they step outside. The issue is not who “owns” particular problems in science, but about developing useful tools for thinking about how things work."

    This keeps coming up, so hopefully will justify a retweet: My argument for why we shouldn't confuse physics as a way of thinking and physics as in "Department of"
    Poništi
  9. 14. pro 2019.

    Nice recent neuroAI research survey by

    2 korisnika označavaju da im se sviđa
    Poništi
  10. 13. pro 2019.

    More generally, replace "bioinformatics" with "mathematical modelling". Any bio department that isn't trying in earnest to navigate this transition should be called out for providing inadequate training.

    We've all heard "bioinformatics is a growth area", but I think the sheer number of job postings at tell us something else: biology is now a quantitative science. We don't need more bioinformaticians. We need to realize that being biologist IS knowing bioinformatics
    2 korisnika označavaju da im se sviđa
    Poništi
  11. 12. pro 2019.

    Trying to navigate the tension: model-based RL feels so right, but compounding errors are so bad...

    Model-Based Reinforcement Learning: Theory and Practice New blog post by about the taxonomy of model-based RL methods, when we should use models, and the state-of-the-art MBPO algorithm for sample-efficient RL.
    2 korisnika označavaju da im se sviđa
    Poništi
  12. proslijedio/la je Tweet
    11. pro 2019.

    Oldies but goldies: Jordan, Kinderlehrer, Otto, The Variational Formulation of the Fokker-Planck Equation. Showed that heat equation is Optimal Transport flow (i.e. flow of particles) minimizing Shannon entropy. Other entropies lead to non-linear PDEs.

    1 reply 64 proslijeđena tweeta 279 korisnika označava da im se sviđa
    Poništi
  13. proslijedio/la je Tweet
    4. pro 2019.

    . & I are co-organizing the next Conference Apr 1, 2020—no joke!—on the topic Triangulating Intelligence: Melding Neuroscience, Psychology, and AI. Botvinick————Tenenbaum——save the date!

    9 replies 73 proslijeđena tweeta 299 korisnika označava da im se sviđa
    Poništi
  14. 28. stu 2019.

    The singular triumph of probably any methodology that distills phenomena into classes dependent on the problem's determining degrees of freedom is the day you assemble the phase diagram for your problem. Today was that day. What a great feeling!

    1 korisnik označava da mu se sviđa
    Poništi
  15. 26. stu 2019.

    is a dying breed in that regard thankfully and you know how they say science advances one "get together" at a time...

    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  16. 26. stu 2019.

    Also issue in caricatured neuroscience, pithily captured (again) by , in description of Neural Networks submitters: not even fig7 worthy. "Relegated" to fig11k no less.

    "Yes"
    Reviewing Editor Raymond Goldstein answers the question: Are theoretical results ‘Results’?
    1 reply 2 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  17. 23. stu 2019.

    Naturalistic behavior in structured environments - modern art style. MTL brews channelling a fav model system for physics of behavior.

    1 reply 4 korisnika označavaju da im se sviđa
    Poništi
  18. proslijedio/la je Tweet
    18. stu 2019.

    Trends in Cognitive Sciences Theories of Error Back-Propagation in the Brain

    1 reply 47 proslijeđenih tweetova 121 korisnik označava da mu se sviđa
    Poništi
  19. 16. stu 2019.

    At . Radoslaw Cichy proposes deepnets as the Duchamp toilet of neuroscience. Not sure I fully get the analogy. But certainly evocative.

    1 proslijeđeni tweet
    Poništi
  20. proslijedio/la je Tweet
    15. stu 2019.

    Montreal AI & Neuroscience MAIN 2019 ! J-1 MAIN lectures (UdeM) J+1 Educational sessions (Mila) Program: Powered by

    , , i još njih 4
    11 proslijeđenih tweetova 15 korisnika označava da im se sviđa
    Poništi

@mptouzel još nije objavio nijedan Tweet.

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:

    ·