Rezultati pretraživanja
  1. 23. velj 2017.

    Congrats "Understanding Deep Learning Requires Rethinking Generalization" for best paper award! Schedule:

  2. 4. svi 2019.

    2/2 Short answer: If you're an algorithm designer, don't quit your day job yet; but don't get too comfortable. Paper: A new dog learns old tricks: RL finds classic optimization algorithms (with William Kong, Chris Liaw, and Aranyak Mehta) (PDF)

    Prikaži ovu nit
  3. 26. tra 2018.

    Our new paper, Hierarchical Density Order Embeddings, is appearing at 2018 (with code)! We learn hierarchical representations of concepts using encapsulation of probability densities.

  4. 23. tra 2019.

    Camera-ready papers for the SafeML workshop are now available! Check them out here:

  5. 23. velj 2017.

    Congrats "Making Neural Programming Architectures Generalize via Recursion" for best paper award! Schedule:

  6. Machine learning excels at finding correlations in data, but causation is a big open research question. At this week, we laid out a new framework for better understanding cause and effect within .

  7. 29. sij 2018.

    YES! "Quantitively Evaluating GANs with divergence proposed for training" got accepted at ICLR. See you soon!

  8. 22. pro 2019.

    Our "Simplified Action Decoder" (, w/ Hengyuan Hu), current SOTA for RL(w/o search) on 2-5 player Hanabi🎇 will be spotlight! The code, , includes trained agents and fast Pytorch version of R2D2&Ape-x 🔥🔥

  9. 1. svi 2018.

    If you like ML and math, and would like to do a postdoc at MILA, reach out to me. If you're at , let's talk in person!

  10. 8. svi 2019.

    How can RL agents learn to learn online? Deep online learning via meta-learning, presented by Anusha Nagabandi and right now at , meta-learning to learn online adaptation for model-based RL with MAML and Bayesian non-parametrics!

  11. 2. velj 2018.
  12. 24. ruj 2019.

    People jumping onto papers last minute to try to grab that coauthorship be like

    Prikaži ovu nit
  13. 19. pro 2019.

    Excited to share that our paper "Semi-Supervised Generative Modeling for Controllable Speech Synthesis" got accepted at   paper: demo:

    Prikaži ovu nit
  14. 23. velj 2017.

    2/3 best papers are from . Neural architecture search by Zoph & Le could have been the 3rd!

  15. 21. pro 2019.

    Have you heard of "RL as Inference"? ... you might be surprised that this framing completely ignores the role of uncertainty! (confusing, since it talks a lot about "posteriors") Our spotlight tries to make sense of this:

  16. 29. sij 2018.

    Moore's law of deep learning: doubles in size every year. (2017: 430 submissions; 2018: 935.)

  17. Happy to share that our paper "Batch-shaping for learning conditional channel gated networks" has been accepted at Paper:  Joint work with &

  18. 6. svi 2019.

    This year's livestream is so much better than previous years, and I can watch workshop talks live 👍👍👍 

    Prikaži ovu nit
  19. 31. sij 2018.
  20. I highly recommend Deep Learning practitioners to read the best paper of 2019. Their proposed algorithm can save you a tremendous amount of time and money on training Neural Networks! Paper:

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.