Tim Herfurth

@tee_8ch

data scientist, physicist, neuroscientist, machine learning researcher, philosophy aficionado; amazed by evolution, science & nature

linkedin.com/in/tim-herfurt…
Vrijeme pridruživanja: ožujak 2015.
0 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @tee_8ch

Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @tee_8ch

  1. proslijedio/la je Tweet
    25. sij

    What GPT-2 accidentally reveals about nature, nurture, and the challenge of building intelligence. Prelude to a longer piece on how to best advance AI over the next decade 😊

    101 proslijeđeni tweet 318 korisnika označava da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    4. pro 2019.

    Our new paper, Deep Learning for Symbolic Mathematics, is now on arXiv We added *a lot* of new results compared to the original submission. With (1/7)

    580 proslijeđenih tweetova 1.725 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet
    22. stu 2019.

    What are Transformer Networks, and why are they getting so much hype? New blog post from Ben Cowley unpacking it for you:

    1 reply 12 proslijeđenih tweetova 66 korisnika označava da im se sviđa
    Poništi
  4. proslijedio/la je Tweet
    28. lis 2019.

    Our opinion piece that deep learning should guide neuroscience research just came out:

    6 replies 175 proslijeđenih tweetova 433 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. 24. lis 2019.

    Dear , in several mindscape episodes you state something like "life converts high-entropy energy from the sun into low-entropy energy". I get the basic idea but from all I know entropy is not a property of energy (or photons). Could help me there?

    Poništi
  6. proslijedio/la je Tweet
    2. lis 2019.

    In new work with , we reconstruct the architecture, weights, and biases of unknown deep ReLU networks by observing their output. ReLU nets are piecewise linear functions - a lot of info is in the boundaries between linear regions. Full paper:

    76 proslijeđenih tweetova 273 korisnika označavaju da im se sviđa
    Poništi
  7. proslijedio/la je Tweet
    4. ruj 2019.

    We are making good progress on , just yesterday I sampled the 8 schools model with NUTS (all in TF, fully compiled and vectorized across chains). Mostly work by . Very hard to say when and if it will supersede , but we will continue to maintain it either way

    . I've assumed PyMC4 will someday replace PyMC3 as the primary PyMC implementation. If that's right, when do you think it's most likely to happen? The repo for 3 has more recent activity than the repo for 4, so I'm wondering if the plans for 4 have changed
    10 proslijeđenih tweetova 68 korisnika označava da im se sviđa
    Poništi
  8. proslijedio/la je Tweet
    21. kol 2019.

    After 2.5 years, we finally shipped the book off for printing. The GitHub graphs do look a bit scary... You can download the PDF (and some jupyter notebooks) from here:

    21 reply 171 proslijeđeni tweet 662 korisnika označavaju da im se sviđa
    Poništi
  9. 21. kol 2019.

    thanks! I think I am mostly interested in DAG-focused tools but I'll try to check all of them. Is R a big thing for those working in causal inference?

    1 reply 1 korisnik označava da mu se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  10. 20. kol 2019.

    is anyone aware of good (and maintained) python tools for causal inference?

    6 replies 5 proslijeđenih tweetova 19 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  11. proslijedio/la je Tweet
    25. tra 2019.

    A decoder, which can transform brain activity into speech, is presented in a study published in Nature. The brain–computer interface device synthesizes speech using the neural signals that control lip, tongue, larynx and jaw movements. Read the paper:

    402 proslijeđena tweeta 753 korisnika označavaju da im se sviđa
    Poništi
  12. proslijedio/la je Tweet
    25. ožu 2019.

    Fun conversation with Sean Carroll on his podcast. More details in a new preprint we have out , and a COSYNE talk at .

    Mindscape Episode 39: Malcolm MacIver on Sensing, Consciousness, and Imagination.
    Prikaži ovu nit
    1 reply 8 proslijeđenih tweetova 14 korisnika označava da im se sviđa
    Poništi
  13. 12. tra 2019.

    excellent intro to causal inference by

    1 proslijeđeni tweet 4 korisnika označavaju da im se sviđa
    Poništi
  14. proslijedio/la je Tweet
    25. ožu 2019.

    1/6 Deep classifiers seem to be extremely invariant to *task-relevant* changes. We can change the content of any ImageNet image, without changing model predictions over the 1000 classes at all. Blog post @ . with Rich Zemel

    209 proslijeđenih tweetova 565 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  15. 24. lis 2018.

    I have to give away my ticket for the meetup Message me name and email and I can reassign my ticket (to first to reply)

    2 proslijeđena tweeta 5 korisnika označava da im se sviđa
    Poništi
  16. 5. pro 2017.

    Opening the black box of deep neural networks via information - part I via

    Poništi
  17. 16. lis 2017.

    Double standards: 'Why aren't we all with Somalia?'

    Poništi
  18. proslijedio/la je Tweet
    9. lis 2017.

    Nice review paper by Tim Herfurth and Tatjana Tchumatchenko in Current Opinion in Neurobiology

    2 proslijeđena tweeta 2 korisnika označavaju da im se sviđa
    Poništi
  19. 22. ruj 2017.

    is there a print version? the link on does not lead me anywhere.

    Odgovor korisniku/ci @tee_8ch
    thanks!
    Poništi
  20. 21. ruj 2017.

    highly recommend A Course in Machine Learning by (see )

    1 reply
    Poništi

@tee_8ch 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:

    ·