Andrej KarpathyOvjeren akaunt

@karpathy

Director of AI at Tesla. Previously a Research Scientist at OpenAI, and CS PhD student at Stanford. I like to train Deep Neural Nets on large datasets.

Stanford
Vrijeme pridruživanja: travanj 2009.

Medijski sadržaj

  1. 3. velj

    Help revolutionize the world with full self-driving by joining us at Tesla Autopilot: It is very hard to find other places where AI expertise makes as much of a difference on as big of a problem.

  2. 15. sij

    Why the Future of Farming is in Cities - The Big Money in Vertical Farming

  3. 4. sij
  4. 4. sij

    Metabolic Engineering and Synthetic Biology of Yeast - Jens Nielsen🤯 (the whole channel is quite good). Bio will grow into a major tech stack. We're writing assembly today, but when the AWS is up things will get interesting.

  5. Still slowly making my way through this year's NeurIPS talks. Esp like to stumble by good talks from slightly different areas, e.g. tonight liked "ML Meets Single-Cell Biology" Incredible that we're mapping out cell state markov chains for tissues.

  6. I implemented some normalizing flows yesterday (NICE, RealNVP, MAF, IAF), tried to make core of it somewhat clean in case helpful I like how flow layers can be structured similar to backprop, each needs an invert() and emits a log det J "regularization"

  7. Age of Empires IV first gameplay footage and a few details released a few days ago!! a lot of my childhood is AoE2 + long overnight "bring your own desktop" LAN parties. *squeal*

    Prikaži ovu nit
  8. I clicked to read an article. Web 2019

  9. Odgovor korisniku/ci

    I love this slide a lot. 👏

  10. Wow, this YouTuber needs his own Netflix TV show. 👏 "Fire Ants vs. Simulated River Jungle", and more generally the Fire Nation's migration to the virgin lands of the Selva de Fuego Paludarium 10/10

  11. "Modern microprocessor built from complementary carbon nanotube transistors" RV16X-NANO, fabricated entirely using carbon nanotube field-effect transistor (CNFET): 16-bit microprocessor based on the RISC-V instruction set. Wafer-scale, VLSI-compatible🔥

  12. Cryogenic electron microscopy (cryo-EM) intro lecture & more detailed class , e.g. allows imaging of different stages of a vicious phage attack on an unsuspecting marine cyanobacteria 😱(from Wei Dai et al 2013):

  13. Autocompletion with deep learning very cool! I tried related ideas a long while ago in days of char-rnn but it wasn't very useful at the time. With new toys (GPT-2) and more focus this may start to work quite well.

  14. Boston Dynamics: New Robots Now Fight Back :O :D

  15. This diagram, which I hastily sketched out in Google slides one day a few years ago at 3am, has become the most popular ugliest diagram of a ConvNet and, I regret to see, the #2 result in Google image search for "ConvNet". I am sorry 😖

  16. Jonathan Blow - Preventing the Collapse of Civilization interesting talk, a bear case for monotonic and exponential progress of technology due to turtles on turtles in our tech stacks and generational loss of knowledge.

  17. The two tech stacks ❤️

  18. An SR latch from while ago that I tried to use for a 555 timer to get clock signal for an all-breadboard transistor scale mini-CPU. Then discovered this is very hard without MOSFETs, but those don’t seem cheaply available as discrete components. And ICs are cheating. Bit stuck 🤷‍♂️

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:

    ·