Visualizing and Understanding Convolutional Networks
deeper layers = more abstract
find images that maximally activate some channels, and also backprop.
model learns in a hierarchical way.
Last Reviewed: 10/25/2025
Visualizing and Understanding Convolutional Networks
deeper layers = more abstract
find images that maximally activate some channels, and also backprop.
model learns in a hierarchical way.
Last Reviewed: 10/25/2025