John Dagdelen

@jmdagdelen

UC Berkeley PhD Student in working at the intersection of materials science, artificial intelligence, and high performance computing.

Berkeley, CA
Vrijeme pridruživanja: kolovoz 2017.

Tweetovi

Blokirali ste korisnika/cu @jmdagdelen

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

  1. Prikvačeni tweet
    20. pro 2019.
    Poništi
  2. 4. velj

    The product viz team at LYNX had way too much fun with their announcement video. I love it.

    Poništi
  3. 4. velj

    The fact that I can super quickly spin up Google Colab + GPU + Google Drive as a filesystem is ML magic to me.

    Poništi
  4. 3. velj

    This should be required reading for new grad students in dev-heavy fields.

    Poništi
  5. 2. velj

    I was at this event and it was scary how much false information speakers like were passing on to their audiences regarding tech and science. For example, she told audience that she met a man who used CRISPR to edit his 1-year-old’s eye color and height.

    Poništi
  6. 2. velj
    Prikaži ovu nit
    Poništi
  7. 2. velj

    Michael Phelps REALLY wants you to know how helpful therapy is.

    Prikaži ovu nit
    Poništi
  8. 30. sij

    I can totally identify. Credit

    Poništi
  9. proslijedio/la je Tweet
    29. sij

    You don't get to write very often about Harvard chemistry professors being hauled off by the FBI. But here we are! Prof. Charles Lieber and China's "Thousand Talents" program: what a mess.

    Poništi
  10. proslijedio/la je Tweet
    27. sij

    Very happy to visit and looking forward to discussions ! Just boarded plane to Chicago 😊

    Poništi
  11. proslijedio/la je Tweet
    26. sij

    Holy crap look at that sweet procedural finger animation based on incoming notes... Or at least that’s how this is done, right? I can’t imagine mocapping every finger motion!

    Prikaži ovu nit
    Poništi
  12. proslijedio/la je Tweet
    26. sij
    Odgovor korisnicima

    I think you're pigeonholing them: they can do many different things with the right training and seed funding.

    Poništi
  13. proslijedio/la je Tweet
    24. sij

    Pigeon-enabled healthcare! I guess it’s still deep learning

    Poništi
  14. proslijedio/la je Tweet
    25. sij

    "Visual Dialogue without Vision or Dialogue" "We develop an embarrassingly simple method (don't look at the picture) that... achieves near state-of-the-art performance" From two years ago but updated. Reminds me of some test-taking strategies! abs:

    Poništi
  15. proslijedio/la je Tweet
    24. sij

    Here's a new ~615/640 nm rhodamine fresh off the prep HPLC. I can't decide which color I like more... the violet color of the solution or its hot pink/magenta fluorescence.

    Poništi
  16. proslijedio/la je Tweet
    13. stu 2019.
    Poništi
  17. proslijedio/la je Tweet
    23. sij

    AI-driven deep rigging?! What if you could use AI to rig and animate a 2D video of a face? 🤯 credits to tpocomp@ on insta

    Prikaži ovu nit
    Poništi
  18. 23. sij

    Always great to rewatch cool ML videos. This is one of my favorites. Deep Fake VFX - Pity the poor impressionist by Jim Meskimen

    Poništi
  19. 23. sij

    Super cool work! Video is a summary of DeepFovea from :

    Poništi
  20. proslijedio/la je Tweet
    21. sij

    Propnet is a knowledge graph connecting materials properties using mostly physics-based models (Cahill thermal cond., Lindemann criterion, Moss relation, etc). Knowledge of some materials properties is used to estimate others. Mrdjenovich et al, Matter

    Poništi
  21. proslijedio/la je Tweet
    21. sij

    Professor Didier de Fontaine's Thermodynamics book makes it amongst World Scientific's 10 best-selling physics books for 2019!

    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:

    ·