Nando de Freitas

@NandoDF

I research intelligence to understand what we are, and to harness it wisely. Principal Scientist at DeepMind, CIFAR Fellow, previously Prof at UBC and Oxford

London, England
Vrijeme pridruživanja: travanj 2009.

Tweetovi

Blokirali ste korisnika/cu @NandoDF

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

  1. 30. sij

    This is brilliant innovative work on machine learning

    Poništi
  2. proslijedio/la je Tweet
    29. sij

    ‘Mapping the future’: our recent paper, which provides insight into previously unexplained elements of dopamine-based learning in the brain, is on the front cover of ! 🎉 Read the blog:

    Poništi
  3. 29. sij
    Poništi
  4. proslijedio/la je Tweet
    28. sij

    Open-domain conversation is an extremely difficult task for ML systems. Meena is a research effort at in this area. It's challenging, but we are making progress towards more fluent and sensible conversations. Nice work, Daniel, & everyone involved!

    Poništi
  5. proslijedio/la je Tweet
    28. sij

    In “Artificial Intelligence, Values and Alignment” DeepMind’s explores approaches to aligning AI with a wide range of human values:

    Poništi
  6. proslijedio/la je Tweet
    23. sij

    Q-learning is difficult to apply when the number of available actions is large. We show that a simple extension based on amortized stochastic search allows Q-learning to scale to high-dimensional discrete, continuous or hybrid action spaces:

    Poništi
  7. proslijedio/la je Tweet
    23. sij

    I'm very pleased to see today's news that the Karsh Family Foundation has made the largest-ever gift to to support STEM education. This will help increase diversity in CS & other STEM fields! Thank you, Bruce and Martha Karsh and !

    Poništi
  8. proslijedio/la je Tweet

    Applications for DL𝕏 SA 2020 close in 10 DAYS!! Why attend? We have a rich lineup of speakers, unconferences and hackathons that will boost your AI/ML ambitions! Already an expert? Come engage with likeminded people and inspire the next generation!

    Poništi
  9. proslijedio/la je Tweet
    21. sij

    We are excited to officially announce details for ! 🌍 This years conference will be held in Accra Ghana🇬🇭, from the 4th to 8th of August, 2020. Mark your calendars and start preparing, we hope to see you all there!💃🏾🥳🎉 Watch this space for more updates!

    , , i još njih 7
    Poništi
  10. proslijedio/la je Tweet
    22. sij

    Excited to share PCGrad, a super simple & effective method for multi-task learning & multi-task RL: project conflicting gradients On Meta-World MT50, PCGrad can solve *2x* more tasks than prior methods w/ Tianhe Yu, S Kumar, Gupta, ,

    Poništi
  11. proslijedio/la je Tweet
    21. sij

    🚨New lecture series🚨 We've teamed up with to bring you the Deep Learning Lecture Series: 12 lectures covering a range of topics in Deep Learning - all led by DeepMind researchers, all free, and all open to everyone. Info & tickets:

    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    20. sij

    Given the smoothness of videos, can we learn models more efficiently than with ? We present Sideways - a step towards a high-throughput, approximate backprop that considers the one-way direction of time and pipelines forward and backward passes.

    Poništi
  13. proslijedio/la je Tweet
    20. sij

    Notice something different? We've got a new and streamlined handle 🙂 Tweet us at from now on! Thank you Twitter! 🙏

    Poništi
  14. proslijedio/la je Tweet
    15. sij

    HO-3D Dataset is now public! It is the first large scale dataset with 3D pose annotations for hands and objects. The annotations are automatically generated by fitting the hand and object models to RGBD sequence obtained from single/multiple cameras. (1/2)

    Prikaži ovu nit
    Poništi
  15. proslijedio/la je Tweet
    15. sij

    🚨🚨Great Internship Opportunity!!! 🚨🚨 Come work with me, , and our awesome multidisciplinary Ethics & Society team as a research intern: Feel free to DM me with any questions!!!

    Prikaži ovu nit
    Poništi
  16. proslijedio/la je Tweet
    16. sij

    Visualization of the air flow field around the nose landing gear of a Boeing 777. The simulation was run on 5,000 Pleiades cores and required over 1 million processor hours, generating over 50 terabytes of data [more: ]

    Poništi
  17. proslijedio/la je Tweet
    16. sij

    If you've wondered - "Which optimizer should I use? ? ? ?" This blogpost by is the best explanation I've seen. It's a surprisingly easy read! Definitely a great / project!

    Poništi
  18. proslijedio/la je Tweet
    15. sij

    The *easiest* way to learn Deep Learning is to build it from scratch! IMO, the same is true when learning a Deep Learning framework. In I show how to build a -like framework Here's the step-by-step code!!

    Poništi
  19. proslijedio/la je Tweet
    16. sij

    How can we predict and control the collective behaviour of artificial agents? Classical game theory isn't much help when there are >2 agents. In our paper, we find markets impose useful structure on interactions between gradient-based learners:

    Poništi
  20. proslijedio/la je Tweet
    16. sij

    Introducing Reformer, an efficiency optimized architecture, based on the Transformer model for language understanding, that can handle context windows of up to 1 million words, all on a single accelerator with only 16GB of memory. Read all about it ↓

    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:

    ·