Mark Sellke

@geoishard

Stanford math grad student.

web.stanford.edu/~msellke/main
Vrijeme pridruživanja: studeni 2016.
0 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @geoishard

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

  1. proslijedio/la je Tweet
    17. sij

    For those who might have missed it, there is a very entertaining thread under this tweet between and that has been going on since November 25th.

    I'm falling in love with random permutations. The average length of the longest cycle in a random permutation of a huge n-element set approaches the "Golomb-Dickman constant" times n. (1/n)
    Prikaži ovu nit
    5 proslijeđenih tweetova 34 korisnika označavaju da im se sviđa
    Poništi
  2. proslijedio/la je Tweet
    15. sij

    Is deep learning is actually performing DEEP learning? We may have given the first proof that neural network is capable of efficient hierarchical learning, while existing theory only shows that deep learning can "simulate" non-hierarchical algorithms

    How does deep learning perform DEEP learning? Microsoft and CMU researchers establish a principle called "backward feature correction" and explain how very deep neural networks can actually perform DEEP hierarchical learning efficiently:
    1 reply 23 proslijeđena tweeta 193 korisnika označavaju da im se sviđa
    Poništi
  3. proslijedio/la je Tweet
    12. sij

    Pluto's moon Nix also rotates chaotically. You could spend a day on Nix where the sun rises in the east and sets in the north! Watch the video to see how weird it is. (3/n)

    37 proslijeđenih tweetova 220 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  4. proslijedio/la je Tweet
    5. sij

    Consider: millions of years ago our antecedents gave a massive sacrifice of their left hemisphere. We lost a tremendous amount of short term memory and replaced it with Broca’s, Wernicke & the phonological loop. But why? So we can—talk. Thus chimpanzees can do this—we can’t:

    945 replies 37.302 proslijeđena tweeta 102.767 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    22. pro 2019.

    I’ve found the master plan of every academic!

    112 replies 2.252 proslijeđena tweeta 8.916 korisnika označava da im se sviđa
    Poništi
  6. proslijedio/la je Tweet
    8. pro 2019.

    Congratulations to our colleague Lin Xiao for the test of time award!!! Online convex optimization and mirror descent for the win!! (As always? :-).)

    1 reply 49 proslijeđenih tweetova 257 korisnika označava da im se sviđa
    Poništi
  7. proslijedio/la je Tweet
    25. stu 2019.

    Mathematician Sarah Peluse has figured out exactly how big a set of numbers needs to be before it's guaranteed to contain a pattern called a "polynomial progression." My story in on an exciting new result in number theory.

    1 reply 47 proslijeđenih tweetova 123 korisnika označavaju da im se sviđa
    Poništi
  8. 5. stu 2019.

    Thrilled to win best paper at SODA 2020. Thanks Seb for the lovely post. It was wonderful to chase convex bodies with you guys.

    First (non-guest) blog post in 5 months (sry!): Convex body chasing, Steiner point, Sellke point, and SODA 2020 best papers.
    1 reply 1 proslijeđeni tweet 30 korisnika označava da im se sviđa
    Poništi
  9. proslijedio/la je Tweet
    21. lis 2019.

    Major progress on the aptly named 'sunflower conjecture' - about how (flower-like) order inevitably arises from randomness. My new story for

    5 replies 48 proslijeđenih tweetova 110 korisnika označava da im se sviđa
    Poništi
  10. 27. ruj 2019.

    Cool paper by Sourav Chatterjee defines an asymmetric correlation coefficient that checks whether Y is a function of X:

    2 proslijeđena tweeta 3 korisnika označavaju da im se sviđa
    Poništi
  11. proslijedio/la je Tweet
    23. ruj 2019.

    gives the evol of inf-width NN; what about finite width? Introducing ** of higher order tensors : each lvl corrects the finite-width error of the prv lvl. Compute ‘em by tensor programs

    1 reply 13 proslijeđenih tweetova 37 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    15. ruj 2019.

    The Randomized Midpoint Method for Log-Concave Sampling.

    1 proslijeđeni tweet 1 korisnik označava da mu se sviđa
    Poništi
  13. proslijedio/la je Tweet
    12. ruj 2019.
    Odgovor korisniku/ci

    Fantastic. My team will reach out. Let’s raise some money for good causes. Looking forward to it. 👍🇺🇸

    819 proslijeđenih tweetova 9.294 korisnika označavaju da im se sviđa
    Poništi
  14. proslijedio/la je Tweet
    6. ruj 2019.

    News! 42 is the sum of three cubes. 42 = (-80538738812075974)^3 + 80435758145817515^3 + 12602123297335631^3

    101 reply 1.799 proslijeđenih tweetova 5.250 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  15. proslijedio/la je Tweet
    26. kol 2019.

    You might be able to train your intuition for this e.g. with this cool little game:

    Odgovor korisniku/ci @webdevMason
    Would be super helpful if every correlation came with a little example graph to show you what it /really/ means
    4 proslijeđena tweeta 12 korisnika označava da im se sviđa
    Poništi
  16. 22. kol 2019.

    Congrats to Ryan Alweiss, Shachar Lovett, Kewen Wu, and Jiapeng Zhang for their progress towards the sunflower conjecture!

    2 korisnika označavaju da im se sviđa
    Poništi
  17. 18. kol 2019.

    Found this cool 2013 article today. NYC charges hundreds of thousands per person per year to sell hotdogs etc at the best locations (in Central Park). But disabled veterans don't have to pay!

    1 korisnik označava da mu se sviđa
    Poništi
  18. proslijedio/la je Tweet
    12. kol 2019.

    What! The sound of cash from an atm is actually fake. The sound tends to reassure humans cash is on the way.

    5 proslijeđenih tweetova 21 korisnik označava da mu se sviđa
    Poništi
  19. 26. srp 2019.

    You can still use old Twitter for now :)

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

    Hao Huang@Emory: Ex.1: ∃edge-signing of n-cube with 2^{n-1} eigs each of +/-sqrt(n) Interlacing=>Any induced subgraph with >2^{n-1} vtcs has max eig >= sqrt(n) Ex.2: In subgraph, max eig <= max valency, even with signs Hence [GL92] the Sensitivity Conj, s(f) >= sqrt(deg(f))

    29 replies 539 proslijeđenih tweetova 1.438 korisnika označava da im se sviđa
    Poništi

@geoishard 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:

    ·