Check out our @googleai blog post on new framework to understand generalization in deep learning. paper: https://t.co/4piRRt2zHI with @bneyshabur, @PreetumNakkiran https://t.co/Wlbk2cH4nz

— Hanie Sedghi (@HanieSedghi) March 10, 2021

Yay! differentiable FFT! https://t.co/BXGOFhFih7

— Yann LeCun (@ylecun) March 10, 2021

We introduce a new approach for image compression: instead of storing the pixels in an image, we store the weights of an MLP overfitted to the image 🌟 At low bit-rates this can do better than JPEG!https://t.co/ATIyOEiwNX

with @adam_golinski @notmilad @yeewhye @ArnaudDoucet1 pic.twitter.com/5sVBc2oST5

— Emilien Dupont (@emidup) March 10, 2021

Today we announce the release of new sparsity features in the #XNNPACK acceleration library that is powering #TensorFlowLite! Sparse inference improves efficiency without degrading quality in applications like Google Meet's background effects. https://t.co/QAcVvSNk5L pic.twitter.com/aYsByJ0Eiz

— Google AI (@GoogleAI) March 9, 2021

This is it. Wonderful comment on how big orgs work. https://t.co/1t418r63MJ pic.twitter.com/ZcTLsPXu7v

— Vicki Boykis (@vboykis) March 9, 2021

🚨Another black box algorithm is being pried open 🚨

Judges have ruled that the makers of a “probablistic genotyping” DNA analysis program used in more than 850 criminal cases must hand over their source code.@lkirchner reports:https://t.co/weP4NBvnOt

— Julia Angwin (@JuliaAngwin) March 9, 2021

A while back, I wrote a Python library for handling YAML-based configuration in my ML projects.

I've been installing (`pip install quinine`) it for my own projects for a while, now you can use it too

README: https://t.co/GNSRvuFp42

— Karan Goel (@krandiash) March 9, 2021

Introducing VISSL (https://t.co/iBEpmCi09R) - a library for reproducible, SOTA self-supervised learning for computer vision! Over 10 methods implemented, 60 pre-trained models, 15 benchmarks, and counting. pic.twitter.com/ZZMd8DpHBD

— PyTorch (@PyTorch) March 9, 2021

This is one of the coolest elevation maps I've seen! Elevation lines of Oahu, #Hawaii. #dataviz

Source: https://t.co/02bmK4aEsR pic.twitter.com/TuWJDNPkT8

— Randy Olson (@randal_olson) March 8, 2021

NEW: we spoke to workers who label data to train AI on platforms like Amazon Mechanical Turk

they're being incentivized to submit biased responses that fall in line with the majority — or risk losing their already precarious work https://t.co/ghqEXBVywf

— The PS1 startup sound, but as a person (@zenalbatross) March 8, 2021

Want to make your results look "incredible"? Turn a modest improvement into a WIRED story? Step 1: run y axis from 62 to 78 (rather than 0 to 100). Step 2: help yourself to an extra billion parameters. Step 3: congratulate yourself.

New SEER paper @facebook pic.twitter.com/y64IXuPy9f

— Gary Marcus (@GaryMarcus) March 7, 2021

First Principles of Computer Vision by Shree Nayar.

In the era of deep learning everything, understanding the fundamentals is more important than ever!https://t.co/wQvdIXC8TM

— Jia-Bin Huang (@jbhuang0604) March 7, 2021