Install Theme

Anyway, while looking over the twitter feed of that “predictive text” guy I found this, which I had never seen before (although @deusvulture says it’s famous)

Recurrent neural networks: what the fuck

(What’s especially impressive is that the parameters were fit using an individual piece of consumer hardware.  High-end consumer hardware, but still.  This isn’t just throwing giant clusters at the problem)

  1. ceruleanvulpine reblogged this from nostalgebraist
  2. sufficientlylargen reblogged this from eikotheblue
  3. vincentmunsmusings reblogged this from wolffyluna
  4. wolffyluna reblogged this from uncrediblehallq
  5. grahamejoyce-blog reblogged this from nostalgebraist
  6. nostalgebraist reblogged this from lithnin and added:
    Yeah, now that you say that it seems right to me – I think I was wrong above.I think there are two effects here pushing...
  7. lithnin reblogged this from nostalgebraist and added:
    It’s funny, when thinking about the same underlying facts I came to the exact opposite conclusions. My expectation is...
  8. adzolotl reblogged this from multiheaded1793
  9. cofinaldestination reblogged this from nostalgebraist
  10. fipindustries reblogged this from nostalgebraist
  11. phenoct reblogged this from nostalgebraist
  12. uncrediblehallq reblogged this from nostalgebraist and added:
    If you look at the Robo Rosewater Twitter account, the cards based on longer training / better input usually end up...
  13. multiheaded1793 reblogged this from multiheaded1793
  14. szhmidty reblogged this from nostalgebraist and added:
    Hold up, depending on the cost that last one seems potentially useful for a black deck. Your opponent summon a legendary...