Simon Osindero

@sindero

Scientist, Engineer, Designer, Artist, Entrepreneur. Researcher at . Personal account, but mostly limited to AI/ML stuff (+ occasional cute animals).

Vrijeme pridruživanja: lipanj 2010.
18 fotografija ili videozapisa Fotografije i videozapisi

Tweetovi

Blokirali ste korisnika/cu @sindero

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

  1. Prikvačeni tweet
    16. pro 2017.

    A few folks asked, and so here’s a transcript of the talk I gave at the workshop dinner at . Thanks again to the organizers and other presenters for a great day!

    70 proslijeđenih tweetova 187 korisnika označava da im se sviđa
    Poništi
  2. prije 12 sati

    👍🏽

    0:45
    In January, , , and I ran a short class at on topics we think are missing in most CS programs — tools we use every day that everyone should know, like bash, git, vim, and tmux. And now the lecture notes and videos are online!
    2 korisnika označavaju da im se sviđa
    Poništi
  3. proslijedio/la je Tweet
    2. velj

    Takes me back. Learned this trick as a teenager and was very impressed with the universe.

    Cut a paper circle, fold it so that the circumference falls on a fixed point. Repeat. This is how you paper-fold an ellipse [source: ]
    4 proslijeđena tweeta 53 korisnika označavaju da im se sviđa
    Poništi
  4. proslijedio/la je Tweet
    1. velj

    I feel like I should pay to run one of these on myself

    I had to get a background check for my job, and it turns out the report is a 300+ page pdf of every single tweet I’ve ever liked with the work “fuck” in it. Enjoy your dystopian bs! *waves*
    Prikaži ovu nit
    13 replies 44 proslijeđena tweeta 570 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  5. proslijedio/la je Tweet
    1. velj

    To the left, you see a trained agent playing a level of a game. To the right, you see the same playthrough from an agent-centric perspective: cropped, translated, and rotated with the agent in the center. Which perspective is the best input for the agent?

    13 replies 32 proslijeđena tweeta 163 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  6. proslijedio/la je Tweet
    30. sij

    For the benefit of all readers, I have compiled my "Causal, Casual and Curious" articles in one searchable page, now posted here: . Enjoy.

    41 proslijeđeni tweet 156 korisnika označava da im se sviđa
    Poništi
  7. 31. sij

    At Harvard there some really great and really talented Profs, but there also some rather mediocre ones who may harbour prejudiced perspectives. Don't let them make you feel like you don't belong...

    today my CS professor verified my imposter syndrome by saying "you're a good student, but here at Harvard there are some really great and really talented students" when he said that memories of loneliness and exclusion within the department flooded my mind, and I started crying
    Prikaži ovu nit
    12 korisnika označava da im se sviđa
    Poništi
  8. proslijedio/la je Tweet
    31. sij

    Still relevant...

    1.980 proslijeđenih tweetova 8.859 korisnika označava da im se sviđa
    Poništi
  9. proslijedio/la je Tweet
    31. sij

    Here is another opportunity to win a travel to present your work at ! In partnership with , is hosting a computer vision for agriculture challenge.

    50 proslijeđenih tweetova 87 korisnika označava da im se sviđa
    Poništi
  10. proslijedio/la je Tweet
    31. sij

    . is the only place crazy enough to conceive of rendering 4K 60+ Hz graphics at home by running a neural network for every frame. DLSS: Trained in , running on the Tensor Cores in RTX GPUs, redefining graphics with AI. And we’re just getting started!

    And just in case you'd forgotten just how BEAUTIFUL looks? You should probably have a quick look at the wonders is pulling off with their DLSS implementation. Its magic. Pure Space Magic.
    5 replies 48 proslijeđenih tweetova 293 korisnika označavaju da im se sviđa
    Poništi
  11. proslijedio/la je Tweet
    31. sij

    RigL accuracy keeps growing and growing, even with a network with very few params! 👇🏿Details below 👇🏿

    How accurate can a ResNet-50 on ImageNet with only 250k of its original 25M weights be🧐? With RigL, accuracy of this 99% sparse network seems to increase with training time almost indefinitely📈! When does it stop?! 1x->53% 🤓 5x->61.8% 😎 10x->63.9% 🤔 40x->66.9% 😵 100x->?
    Prikaži ovu nit
    2 proslijeđena tweeta 10 korisnika označava da im se sviđa
    Poništi
  12. 31. sij

    Github link:

    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  13. 31. sij

    I've only taken a quick look so far, but this seems like a really great framework! Looks like it might be particularly useful for folks who are getting started and/or who want to focus on the model & algo, without worrying about lots of the other moving pieces.

    Catalyst – one year after open-sourcing, Today we finally reached 1k ⭐️ on ! Thanks to all our contributors, users and kagglers :) Catalyst – reproducible and fast DL & RL. Powered by . 🚀
    Prikaži ovu nit
    1 reply 4 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  14. proslijedio/la je Tweet
    31. sij

    This is a nice package for making pyplot animations more intuitive: All you do is call "camera.snap()" every time you re-do the plot.

    141 proslijeđeni tweet 562 korisnika označavaju da im se sviđa
    Poništi
  15. proslijedio/la je Tweet
    31. sij

    The Golden Gate Stem Fair is the Bay Area's regional high school science fair. It touches the lives of thousands of local kids and is at risk this year (again) of not happening due to lack of funds. Can we raise $15k to save it? (retweets appreciated!)

    8 replies 66 proslijeđenih tweetova 159 korisnika označava da im se sviđa
    Poništi
  16. proslijedio/la je Tweet
    31. sij

    This is one of the most convincing attacks on the US healthcare system I’ve read, mainly because it’s an attempt to defend it.

    THREAD: Let's examine all the possible ways this tweet represents -- and one of the sillier arguments for I've heard in a long time... 1/
    Prikaži ovu nit
    5 replies 30 proslijeđenih tweetova 88 korisnika označava da im se sviđa
    Poništi
  17. proslijedio/la je Tweet
    30. sij

    Kenny's fabulous discovery of cycling between 2 trajectories at the choice point is now live at . Go check it out!

    Constant Sub-second Cycling between Representations of Possible Futures in the Hippocampus
    12 proslijeđenih tweetova 24 korisnika označavaju da im se sviđa
    Poništi
  18. proslijedio/la je Tweet
    30. sij

    Whenever I see someone on one of these, I like to imagine they started out with a much larger vehicle and their journey has been riddled with mishaps.

    569 replies 19.211 proslijeđenih tweetova 139.269 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  19. 30. sij

    Of course, it's also interesting (and sometimes necessary) to push what can be done "from scratch" / with minimal prior knowledge as well.

    2 korisnika označavaju da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  20. 30. sij

    I'm a big believer in leveraging curricula, demonstration data, learning from other pre-trained agents, co-evolution, and so on -- especially for hard RL problems.

    Humans learn from curriculum since birth. We can learn complicated math problems because we have accumulated enough prior knowledge. This could be true for training a ML/RL model as well. Let see how curriculum can help an RL agent learn:
    1 reply 11 korisnika označava da im se sviđa
    Prikaži ovu nit
    Prikaži ovu nit
    Poništi
  21. proslijedio/la je Tweet
    30. sij

    We're standardizing OpenAI's deep learning framework on PyTorch to increase our research productivity at scale on GPUs (and have just released a PyTorch version of Spinning Up in Deep RL):

    25 replies 549 proslijeđenih tweetova 1.920 korisnika označava da im se sviđa
    Poništi

@sindero još nije objavio nijedan Tweet.

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:

    ·