Ben Bartlett

@bencbartlett

Physics PhD student , previously • studies nanophotonics ✨ and quantum computers ⚛️, physics, memes, cats • keywords: 🌴🤖🐈🏳️‍🌈

Vrijeme pridruživanja: siječanj 2012.

Tweetovi

Blokirali ste korisnika/cu @bencbartlett

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

  1. Prikvačeni tweet
    23. lis 2019.

    My new paper on photonic quantum programmable gate arrays is now on arXiv! 🍾⚛️🎉 We describe an architecture for a nanophotonic integrated circuit which can be reprogrammed to perform any quantum computation.

    Prikaži ovu nit
    Poništi
  2. proslijedio/la je Tweet
    2. velj

    I used to wonder why the interfaces on Star Trek are so clunky given that it's centuries in the future, but I guess that's just Enterprise software for you

    Prikaži ovu nit
    Poništi
  3. proslijedio/la je Tweet

    how to mess with your coworkers

    Poništi
  4. proslijedio/la je Tweet
    31. sij

    Like a pebble in a stream. Waves, vortices, turbulence.

    Prikaži ovu nit
    Poništi
  5. 30. sij

    and the last two years should just be playing with a charged slinky while moving at .99c

    Prikaži ovu nit
    Poništi
  6. 30. sij

    the first two years of undergrad physics should just be playing with a slinky 5 hr a day

    Prikaži ovu nit
    Poništi
  7. 29. sij

    The animation in the first tweet was likely generated by taking a starting input vector and uniformly rotating it around in a high dimensional circle, which smoothly varies the images output by the generator network and loops the animation back to the original image. [5/5]

    Prikaži ovu nit
    Poništi
  8. 29. sij

    You train both networks and if one doesn't overpower the other (which can be tricky in practice), then eventually the discriminator becomes very good at identifying real images and the generator gets better at fooling the discriminator with very realistic images. [4/5]

    Prikaži ovu nit
    Poništi
  9. 29. sij

    The generator tries to transform a random high-dimensional input vector into a vector that represents a sample from a distribution, such as pictures of landscapes. The discriminator takes in both real and generated images and assigns probabilities that an image is fake. [3/5]

    Prikaži ovu nit
    Poništi
  10. 29. sij

    GANs were first proposed by et al in this 2014 paper: A GAN is a system of two competing neural networks - a "generator" and "discriminator" - which are trained against each other (hence "adversarial" learning). [2/5]

    Prikaži ovu nit
    Poništi
  11. 29. sij

    This type of trippy video is generated by stepping through the latent space of a machine learning model called a generative adversarial network (GAN). Here's a short thread on how they work: [1/5] (Video source: )

    Prikaži ovu nit
    Poništi
  12. 28. sij

    and if you buy the newest version of Pokemon then you can evolve your Σ into whatever Generation VIII bullshit this is:

    Prikaži ovu nit
    Poništi
  13. 28. sij

    if you walk your Σ a along a closed path ∂S then you can evolve it into a:

    Prikaži ovu nit
    Poništi
  14. 28. sij

    other obscure evolutions: if you give your Σ a 𝘃𝗼𝗹𝘂𝗺𝗲 𝘀𝘁𝗼𝗻𝗲 it will evolve into:

    Prikaži ovu nit
    Poništi
  15. 28. sij

    +, Σ, and ∫ are just different evolutions of the same Pokémon

    Prikaži ovu nit
    Poništi
  16. 26. sij

    The Leidenfrost effect creates a layer of vapor below the water and makes its motion nearly frictionless

    Poništi
  17. proslijedio/la je Tweet
    8. stu 2019.

    Visualizing the projection of a rotating tesseract is less confusing if you compare it with the projection of a rotating cube

    Prikaži ovu nit
    Poništi
  18. proslijedio/la je Tweet

    comprehensive diagram of the salient layers of modern computer architecture

    cache
cache
cache
cache
cache
cache
cache
cache
cache
cache
cache
cache
    Prikaži ovu nit
    Poništi
  19. 24. sij

    I am become decoherence, the destroyer of computations

    Poništi
  20. 21. sij

    When you burn steel wool, it gets heavier! While burning, the steel wool oxidizes to form iron oxide: 2Fe + 3O₂ → 2Fe₂O₃

    Poništi
  21. 19. sij

    cop: do you know why I pulled you over? me: I swear that light was green cop: you were doing 450,000,000 in a 40

    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:

    ·