Dileep George

@dileeplearning

Founder at Vicarious AI ( ). Previously Founder and CTO at Numenta. Triply EE (BTech IIT-Mumbai, MS&PhD Stanford)

Vrijeme pridruživanja: lipanj 2017.

Tweetovi

Blokirali ste korisnika/cu @dileeplearning

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

  1. Prikvačeni tweet
    22. ožu 2019.

    Evolution, inductive biases, sample complexity, neocortex, and AGI are all connected in an interesting way. Checkout my HLAI talk.

    Poništi
  2. proslijedio/la je Tweet
    prije 22 sata

    If u ever wondered about the inner workings of 's RCNs, I just published an article on it:

    Poništi
  3. prije 18 sati

    This blog on RCN model is just excellent with amazing visuals! I highly recommend it, especially if you are interested in generative visual models based on the visual cortex.

    Poništi
  4. proslijedio/la je Tweet
    29. sij
    Odgovor korisniku/ci

    Nice video. This is exactly the innateness that is used in our paper "Different clones for different contexts" () and in our cognitive maps paper that learns spatial layout from from ordinal sequential info

    Poništi
  5. proslijedio/la je Tweet
    28. sij
    Odgovor korisniku/ci

    Why do you consider that image interpretation be non-compositional? It is context-sensitive and compositional , and we did tackle that problem in our RCN paper. See Fig 1C(i) where interpretation switches between 'm' and 'un', and (iii) 'bison' and 'bike'

    Poništi
  6. 25. sij

    no, but my kids do the unpacking, looking for toys.

    Poništi
  7. proslijedio/la je Tweet
    21. sij
    Odgovor korisnicima

    The irony is that "flapping is not required to fly" was something Wright brothers learned from observing birds! They kept detailed notes about birds soaring for long distance with no movement of their wings, except for the slight twist of ends for control.

    Poništi
  8. proslijedio/la je Tweet
    15. sij
    Odgovor korisnicima

    I disagree. Effectively you are trying to do what evolution did. It is much more efficient to learn from what evolution created. If you have AGI, you won't ask AGI to be dumb and recreate evolution. Humans are AGI -- we can do better than just starting from primordial soup.

    Poništi
  9. 8. sij

    I will be speaking at the 'Brain Inspired Computing Congress' in silicon valley (April 21-22) and sharing some exciting new work on cortical microcircuits. (DM me for discounted registration code.)

    Poništi
  10. 7. sij

    Life if a vector, brain is a computer, everything is a Markov chain..: let's keep abstracting till we find something we can't disagree on.. :)

    Poništi
  11. proslijedio/la je Tweet
    4. sij

    As to why I feel the timing is ripe? It's been a common thread of work from , part of the fresh "upside-down RL" from , and very recently appearing from .

    Prikaži ovu nit
    Poništi
  12. 4. sij

    Very nice blog. I think neocortex uses a dozen or so general-purpose priors that help with learning and inference, and understanding how to encode those priors is an important challenge for AI...as I put it here:

    Poništi
  13. proslijedio/la je Tweet
    2. sij

    1/ Reflecting back on our DL/CI discussion with , , etal, we should note that CI folks are invoking the same scientific argument that has brought DL its fame and success. Back propagation and subsequently gradient-based optimization were inspired by

    Prikaži ovu nit
    Poništi
  14. 2. sij

    Yoshua wants a new term for 'specific forms of deep learning', (which includes causality, reasoning etc?) Looks like he has almost spelled it out: Deep Learning 2.0

    Poništi
  15. 2. sij
    Poništi
  16. proslijedio/la je Tweet
    1. sij

    Just published my 2nd of 32 annual reviews of my January 1st 2018 hype dampening predictions about self driving cars, AI & ML, and the space industry. My predictions had dates attached so you can evaluate how well I did.

    Poništi
  17. 1. sij

    wishes you a Happy New Year, really!

    Poništi
  18. 31. pro 2019.

    We need a green new deal and a green new DL!

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

    ┏━━┓┏━━┓┏━━┓┏━━┓ ┗━┓┃┃┏┓┃┗━┓┃┃┏┓┃ ┏━┛┃┃┃┃┃┏━┛┃┃┃┃┃ FOCUS SOLELY ON SUBJECT NOT THE SCIENTIST, IN A DEBATE ┃┏━┛┃┃┃┃┃┏━┛┃┃┃┃ ┃┗━┓┃┗┛┃┃┗━┓┃┗┛┃ ┗━━┛┗━━┛┗━━┛┗━━

    Prikaži ovu nit
    Poništi
  20. proslijedio/la je Tweet
    28. pro 2019.

    It was really fun talking with . He asks terrifically insightful questions!

    Poništi
  21. proslijedio/la je Tweet
    28. pro 2019.
    Odgovor korisniku/ci

    those who are interested in this might want to check out our work on ‘cognitive programs’ where abstract concepts are learned as programs on a ‘visual cognitive computer’. And yeah, it transfers between tasks.

    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:

    ·