Live data from Hacker News

Viewing profile — ajtulloch

ajtulloch

HN member
Joined
Sun, Apr 01, 2012, 9:55 PM UTC
HN karma
1,055
Public activity
241 items

About ajtulloch

Andrew Tulloch. Member of Technical Staff at OpenAI. Formerly distinguished engineer at Facebook/Meta.

Find me at http://tullo.ch and github.com/ajtulloch.

Email is andrew@tullo.ch

Recent public activity

  1. comment
  2. comment
    Comment #39890694

    - https://www.cs.utexas.edu/users/flame/laff/pfhp/index.html (e.g. here https://www.cs.utexas.edu/users/flame/laff/pfhp/week2-blocki... ) - https://gist.github.com/nadavrot/5b35d44…

  3. comment
    Comment #38172130

    It's quite unfavorable on modern hardware. A Sapphire Rapids core can do 2 separate 32 half-precision FMAs (vfmadd132ph, [1]) per clock, which is 128 FLOPs/cycle. It is not possibl…

  4. comment
    Comment #37591981

    There’s a big long list of these in https://web.math.princeton.edu/generals/ , not just Tao’s.

  5. comment
    Comment #37356100

    I think the mathematical concept that you are looking for is that of the dual space. Essentially if you have a vector space V, you can construct a dual space V* where the elements …

  6. comment
    Comment #36928688

    One way to view this formula is to use the fact that the Beta distribution is a conjugate prior for the binomial distribution. Essentially if you have a Beta(a, b) prior then your …

  7. comment
    Comment #36759197

    https://www.inference.org.uk/itprnn/book.pdf is a classic text on this connection.

  8. comment
    Comment #36452411

    You're probably thinking of https://en.wikipedia.org/wiki/Oliver_Cromwell and not https://en.wikipedia.org/wiki/Thomas_Cromwell .

  9. comment
    Comment #35464603

    https://github.com/facebook/folly/blob/main/folly/docs/Synch...

  10. comment
  11. comment
    Comment #34785267

    https://quantian.substack.com/p/from-trinity-to-liquidity is an excellent example of applying this idea to the problem of inferring the yield of a nuclear explosion and price impac…

  12. comment
    Comment #33261616

    “Fast Search in Hamming Space with Multi-Index Hashing” ( https://www.cs.toronto.edu/~norouzi/research/papers/multi_in... ) is a great paper. Note that you can do significantly bet…

  13. comment
  14. comment
    Comment #31734419

    TF32 is not IEEE-754 float32, it is a reduced precision format designed for machine learning usecases. The correct specsheet number for FP32 (and FP64 which is the relevant precisi…

  15. comment
    Comment #31464934

    https://news.ycombinator.com/item?id=7266618

  16. comment
    Comment #30057934

    The GotoBLAS paper (“Anatomy of High-Performance Matrix Multiplication”, https://www.cs.utexas.edu/users/flame/pubs/GotoTOMS_final.pd... ) is really a masterpiece. It’s worth inter…

  17. comment
    Comment #29945120

    [flagged]

  18. comment
    Comment #29860899

    I think the two key texts here are Graeber’s Bullshit Jobs and Mark Fisher’s Capitalist Realism, specifically his riff on “market Stalinism”: > The idealized market was supposed to…

  19. comment
    Comment #29520675

    If you see his job to be at the cutting edge of Ruy/Marshall/anti-Marshall the results of games 1-6 suggest he’s done a phenomenal job.

  20. comment
    Comment #29247342

    https://rigtorp.se/notes/hashing/ and links therein ( https://github.com/martinus/better-faster-stronger-mixer ) are almost certainly preferable to this mixer.

  21. comment
    Comment #29234646

    Good news! https://pytorch.org/docs/stable/notes/hip.html

  22. comment
    Comment #28881625

    and he's a genius!

  23. comment
    Comment #28740744

    For folks wondering about applications of this theorem: it is a key building block in the theory of reproducing kernel Hilbert spaces (RKHS), which in turn are the building block o…

  24. comment
    Comment #28300528

    Underappreciated fact - the Apple SoCs since A10 (iPhone 7, etc) are also big.LITTLE designs! https://en.m.wikipedia.org/wiki/Apple_A10

  25. comment
    Comment #26169080

    A bit petty, but the first example is an unstable softmax implementation: defn softmax(t) do Nx.exp(t) / Nx.sum(Nx.exp(t)) end See https://ogunlao.github.io/2020/04/26/you_dont_rea…