Twitter | Pretraživanje | |
Niru Maheswaranathan
Trying to separate signal from the noise; research engineer at Google Brain. Formerly , , . ’s +1. Opinions my own. ⚽️☕️👨🏾‍💻
1.692
Tweetovi
783
Pratim
1.647
Osobe koje vas prate
Tweetovi
Niru Maheswaranathan proslijedio/la je tweet
Adam J Calhoun 18. pro
Peering into a deep network trained on retina data: - Instantaneous RFs are context-dependent and state-based - Network subspace used by white noise and natural scenes are different [Nice update from ]
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 10. pro
Odgovor korisniku/ci @RmHoora
Thank you for the kind words!
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Alex Williams🌹 10. pro
Co-presenting this work with , , , now at poster #156!
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 1. pro
Awesome work, congrats ☺️
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Hidenori Tanaka 21. stu
New paper out on : “From deep learning to mechanistic understanding in neuroscience: the structure of retinal prediction” with fantastic collaborators , , , Stephen Baccus, .
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Dr Lila Landowski 14. stu
Here you can see two neurons sensing one another and connecting in a petri dish. There are 86 billion neurons in the , and they use these webbed hand like structures (“growth cones”) to search for and connect to other neurons or body parts as we develop
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
David Duvenaud 11. stu
We approximated the implicit function theorem to tune millions of hyperparameters. Now we can train data augmentation networks from scratch using gradients from the validation loss. With and
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Jessica AF Thompson🧠🤖🤔👩‍💻✊ 7. stu
So grateful for 's colab on CKA. It seems obvious in retrospect but I hadn't considered the equivalence of calculating similarities based on examples and based on features. My experiments are so much faster now...🚀
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 7. stu
Loved listening to this interview with Andrew Saxe. Great summaries of a lot of beautiful work! (side note, the previous interviews are just as good--kudos to for running a great podcast!)
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Shahab Bakhtiari 2. stu
No, this isn’t from et al recent perspective on . This is David Robinson trying to make a similar point in 1992:
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 1. stu
submission deadline: 31Oct 11:59 **Pacific Time**. registration deadline: 31Jan 11:59 **Eastern Time**. ()
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 1. stu
itermplot () is a matplotlib backend that displays directly in your terminal (iterm2). Really awesome resource if you (like me) enjoy working directly from the IPython repl!
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
srdjan ostojic 1. stu
We received ~650 abstracts for , a number comparable to two years ago in Denver (700), and a big drop wrt Lisbon last year (1000).
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan proslijedio/la je tweet
Brandon Amos 28. lis
Stoked to share a milestone project for all of us! paper with , , S. Boyd, S. Diamond, : Differentiable Convex Optimization Layers Paper: Blog Post: Repo:
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 18. lis
Odgovor korisniku/ci @arunchaganty @EloquentLabs i 2 ostali
Congrats Arun!
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 6. lis
Odgovor korisniku/ci @SuryaGanguli @Stanford @GoogleAI
You’re right, the lab has come a long way since those early days! It’s been a ton of fun watching it grow over the years :)
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 5. lis
Odgovor korisniku/ci @SuryaGanguli @Stanford
Congrats, Surya!!!
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 21. ruj
Odgovor korisniku/ci @sarahookr @lisacrost
Interesting! I feel like the use of a diverging colormap is a poor choice—by changing the range used to center the data, you could tell a different story. This article ( ) by is not about colormaps, but discusses misleading climate visualizations
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 9. ruj
Odgovor korisniku/ci @niru_m
s/algorithms/developments. Matlab is not an algorithm 😅
Reply Retweet Označi sa "sviđa mi se"
Niru Maheswaranathan 9. ruj
The greatest numerical algorithms of the 20th century, according to Nick Trefethen circa 2005 ()
Reply Retweet Označi sa "sviđa mi se"