adam-oberman comments on Deep sparse autoencoders yield interpretable features too

[adam-oberman]

Interesting! Thanks for the post!