[Saxe][Choe] Andrew Saxe on analyzing simple neural networks to understand neural networks better

Feb 11, 2021

A very interesting talk by Andrew Saxe, showing how experiments with really simple neural networks can help us better understand what’s going on in these neural networks and how to improve them.

Yesterday, I attended a symposium (https://sites.google.com/mila.quebec/senai/home)) where he gave a similar talk and I found his work and his approach to be very interesting.

https://www.youtube.com/watch?v=C0PmJZR5rFE

YOUTUBE.COM

Training and generalization dynamics in simple deep networks

https://www.youtube.com/watch?v=C0PmJZR5rFE

← Back to all articles         Quick Navigation:    Next:[ j ] – Prev:[ k ] – List:[ l ]