Tom Brown

@nottombrown

AI research . Previously breaking neural nets .

SF
Vrijeme pridruživanja: lipanj 2009.

Tweetovi

Blokirali ste korisnika/cu @nottombrown

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

  1. Prikvačeni tweet
    13. ruj 2018.

    Astonishingly, no known machine learning system can reliably tell a *bird* from a *bicycle* when there’s an adversary involved. My colleagues and I have proposed a contest to see if we can change this.

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Prikaži ovu nit
    Poništi
  2. 14. sij

    (Photo above is "The Theatre of Disappearance” by Adrián Villar Rojas)

    Prikaži ovu nit
    Poništi
  3. 14. sij

    For the last few months, has been turning my phone into a fancy 3D scanner. Now lets you edit scans too I have a strange love of remixed classical sculptures, and this is one step closer to making my own. HT

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Prikaži ovu nit
    Poništi
  4. 27. pro 2019.

    I now suspect that I have worked with several spies

    Prikaži ovu nit
    Poništi
  5. 27. pro 2019.

    In 1944, the CIA distributed a Simple Sabotage Field Manual with instructions for discretely making organizations less effective. Useful lens for noticing where organizations can become unintentionally self-sabotaging. HT

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Prikaži ovu nit
    Poništi
  6. 5. pro 2019.

    OpenAI is hiring a Head of Security: This is one of the most important hires for OpenAI to make correctly if we want to succeed in having a positive impact on the world. Ping me if you know of folks that might be a good fit - I'm tom@openai.com

    Poništi
  7. proslijedio/la je Tweet
    21. lis 2019.

    Excited to share our work: collaboration requires understanding! In Overcooked, self-play doesn't gel with humans: it expects them to play like itself. (1/4) Demo: Blog: Paper: Code:

    Prikaži ovu nit
    Poništi
  8. 23. stu 2019.

    I was lucky enough to work with Colin as I was starting out in ML research. For anyone who's interested in doing a PhD , I highly recommend talking with him. Congrats, Prof Raffel!

    Poništi
  9. 15. lis 2019.

    No giraffes were in the training set for this hand. The model was just trained on TONS of random environments, so instead of memorizing solutions to specific envs, the easiest things for the model to do was to figure out a general strategy (which included giraffe resistance!).

    Poništi
  10. 15. lis 2019.

    This stuffed-giraffe-resistant hand from my roboticist friends at brings joy to my heart. Problem: Deep learning has trouble generalizing outside the training set. Solution: Put WAY MORE STUFF in the training set.

    Poništi
  11. 4. lis 2019.

    Stuart Russell points to fossil-fuel mega-corporations as examples of powerful and misaligned entities that we built. Corporations were designed to maximize shareholder value, but in this case they worked to outsmart and manipulate the rest of humanity.

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Poništi
  12. 4. lis 2019.

    Yoshua Bengio working hard to moderate an impromptu debate on AI value alignment on 's FB wall. Fascinating thread.

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Poništi
  13. proslijedio/la je Tweet
    18. ruj 2019.

    I couldn't be more excited to be running the AI Fellowship program for the 3rd year - it's my primary priority in my work at and I'm very passionate about it! If you have any Qs, please just ask! Many of the current fellows are also on twitter, and very friendly :)

    Poništi
  14. 15. ruj 2019.

    This lit review from points out three core questions in progress studies. Fields can be defined by their questions or by the tools used to answer questions. Curious what tools will be developed to answer these questions on progress.

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Poništi
  15. 22. kol 2019.

    Also a reminder of my favorite definition of an adversarial example, via : "Adversarial examples are inputs to machine learning models that an attacker has intentionally designed to cause the model to make a mistake."

    Prikaži ovu nit
    Poništi
  16. 22. kol 2019.

    This builds on a long line of research into the question of "what sort of adversarial robustness should we care about?", and we owe a large debt to and for their work on this topic. 2/

    Prikaži ovu nit
    Poništi
  17. 22. kol 2019.

    Most adversarial robustness work focuses on the L_inf and L_2 perturbation attacks (red box in the grid below). The new UAR metric lets researchers test their defenses on a much wider range of attacks and see which defenses give broad benefits and which do not. 1/

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Prikaži ovu nit
    Poništi
  18. 1. kol 2019.

    Learning Day fills my heart with green fluorescent joy. 💚📗🍏💚

    Ovo je potencijalno osjetljiv multimedijski sadržaj. Saznajte više
    Poništi
  19. 1. kol 2019.

    Learning Day! Today I'll be learning about GPU kernel programming by going through the Numba tutorials and writing some CUDA kernels from scratch. 🍿🌽🌰 <= my kernels

    Poništi
  20. 1. kol 2019.

    Learning Day rules!! 📚🍎 People post what they're learning in Slack, and if there are others who are interested then we can learn together. We often learn faster together because each person knows a different piece of the full story.

    Poništi
  21. 30. srp 2019.
    Prikaži ovu nit
    Poništi

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:

    ·