Steven Hansen

@Zergylord

Research Scientist at DeepMind. Slowly learning to learn fast. All opinions my own?

London, England
Vrijeme pridruživanja: lipanj 2009.

Tweetovi

Blokirali ste korisnika/cu @Zergylord

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

  1. Prikvačeni tweet
    30. lis 2019.

    Excited to announce our work on memory generalization in Deep RL is out now! We created a suite of 13 tasks with variants to test interpolation and extrapolation. Our new MRA agent out-performs baselines, but these tasks remain an open challenge. 1/n

    Prikaži ovu nit
    Poništi
  2. 31. sij

    There is so much potential here: That feeling when you've been in the saddle for too long. Drinking the BLEUs away. Something something Lasso regression... Hell, "Long short-term memories" is even a good album title.

    Poništi
  3. 30. sij

    Fair point, but it's worth noting that there is a trade-off between # seeds, # baselines, and # and complexity of environment(s). Personally, I'd prefer a method /w 3 seeds eval'd on 57 tasks with 5 baselines to one with 10 seeds eval'd on 1 task with 2 baselines.

    Poništi
  4. 10. sij
    Poništi
  5. 29. pro 2019.

    Aristotelian physics may be flawed, but it has been unfairly discarded in favor of Newtonian physics, despite the latter not solving all of the open problems in the field. Clearly the way forward is a hybrid system, wherein aether obeys the conservation of momentum.

    Poništi
  6. 27. pro 2019.

    Episodic coverage-based exploration is a great idea, and this instantiation of it yields great results on hard exploration Atari games. Great work from my colleagues at

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

    Happy to have worked with on research combining behavioural mutual information and successor features, which has been accepted for oral presentation at ICLR. Favorite part: clean answer to where to get the “features” for successor features.

    Prikaži ovu nit
    Poništi
  8. 19. pro 2019.

    Conference-decision-anticipation at an all time high. I need that sweet sweet email notification!

    Poništi
  9. 16. pro 2019.

    I know people who think the original meme was "all your Bayes are belong to us". Can't tell if that makes them more or less nerdy 🤔

    Poništi
  10. proslijedio/la je Tweet
    13. pro 2019.

    We just released our scientific analysis of OpenAI Five: We are already using findings from Five in other systems at OpenAI like Dactyl () or our multi-agent work (). Hope that others find the results useful!

    Poništi
  11. proslijedio/la je Tweet
    12. pro 2019.

    Very proud of this work. I am not at NeurIPS this year, but my awesome co-authors Melissa Tan, and are presenting the poster (#192). 🙃🙂🙃

    Poništi
  12. 12. pro 2019.

    Very proud to be a part of this! The docker release makes installing and using these environment painless. Test it out for yourself and come yell at me if something goes wrong ;)

    Poništi
  13. 11. pro 2019.

    Is it too late to rebrand entropy regularization as "the free-will prior"?

    Poništi
  14. 11. pro 2019.

    "now let's talk about consciousness" Can we not?

    Poništi
  15. 9. pro 2019.

    I like London, but it'll be so nice to just forget about Brexit and think about nothing but AI for a week. Day 1:

    Poništi
  16. 9. pro 2019.

    I'm at all week; DMs always open so let me know if any AI Twitter folk wanna chat about intrinsic motivation, generalization in DRL, and/or AI Twitter drama :) And do come by our wonderful poster on Thursday to check out how well memory-based DRL agents generalize!

    Poništi
  17. 9. pro 2019.

    Great tutorial at Turns out that "dataset shit" is really hard.

    Poništi
  18. 8. pro 2019.

    Lightning in Palo Alto? What a narrative violation.

    Poništi
  19. proslijedio/la je Tweet
    7. pro 2019.

    May be I am the only weird one who feels uncomfortable with the idea that one of the pre-readings list includes chapters from three books that are not publicly available as free versions. Perhaps can share those chapters with us? Or is it indirect books marketing?

    Poništi
  20. 30. stu 2019.

    Unblocked, so ignore that first bit ;) Still curious about the connections between reversible networks and Gary's ideas though!

    Prikaži ovu nit
    Poništi
  21. 30. stu 2019.

    Got blocked by for not sufficiently appreciating my time with his 2001 book, The Algebraic Mind. Re: universally qualified one-to-one mapping, iRevNets satisfy this property, right? Would be curious to know what he has to say on that paper:

    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:

    ·