David Sussillo   

@SussilloDavid

Computational neuroscience / deep learning. Scientist at Google Brain, adjunct prof at Stanford. All my tweet are belong to me. 😋🍑🍃

Portola Valley, CA, USA
Vrijeme pridruživanja: studeni 2016.

Medijski sadržaj

  1. Odgovor korisniku/ci

    Me thinking about brains and artificial neural networks.

  2. Odgovor korisniku/ci

    True also for Californians, surprisingly.

  3. JAX + TPUs + Colab

    Prikaži ovu nit
  4. JAX + TPUs + Colab

    Prikaži ovu nit
  5. JAX + TPUs + Colab

    Prikaži ovu nit
  6. JAX + TPUs + Colab

    Prikaži ovu nit
  7. An extremely creative neighbor made Shelob out of produce and branches for Halloween! 🕷 🎃

  8. NeurIPS submissions 🤣

  9. Searching through some of my old academic code from my pre-Google days and found this gem: assert B - 7 // 8 % 0 = 17, “This business will get out of control. It will get out of control and we'll be lucky to live through it.”

    Prikaži ovu nit
  10. Odgovor korisniku/ci
  11. Odgovor korisnicima i sljedećem broju korisnika:

    Here's a slide of RNN studies mostly + some DNN studies. It's a talk slide so I include some of my own work (tacky? but better than ghosting). Also noticeably missing is Jazayeri lab work.

  12. Odgovor korisnicima
  13. Odgovor korisnicima

    I can't code without my radio!

  14. Odgovor korisniku/ci
  15. Meanwhile, my colleagues are now, after reading this story, like, “Daaaaaaamm, Sussillo can’t derive a Kalman filter!” 🤣😂😅

    Prikaži ovu nit
  16. And that was basically that. Some rando accused me academic fraud because I couldn't derive the Kalman filter in real time during a pop quiz in an interview. My only regret is that I didn't angry. The End, thanks for reading. 😎

    Prikaži ovu nit
  17. He asked me another technical question or two, and I went through the motions on board, but I my heart was no longer in it. Finally, I had the presence of mind to say, "I don't think this interview is working out. I'm leaving." And I left the interview and the office.

    Prikaži ovu nit
  18. The interviewer says to me, and this is nearly a verbatim quote, as it's been burned into my skull, "I do not see how you could honestly come by first authorship on those Nature papers if you cannot derive a Kalman filter."

    Prikaži ovu nit
  19. The Germans do so love their titles.😆

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:

    ·