how to extract (plot) the hidden units/softmax (features) from a autoencoder using Tensorflow

Asked Feb 10 '16 at 12:10

Active Feb 10 '16 at 12:28

Viewed 348 times

I'm new with ML, and I'm using tensorflow.
I want to see the features of my autoencoder, but I don't know how to extract(see) the hidden units.
could someone help me?

I made my own dataset, but the original code is with MNIST: https://jmetzen.github.io/2015-11-27/vae.html

I added a layer of hidden units to transform it to Stacked AE.. (I don't know if it became stacked, but worked)

my AE: https://github.com/Snowedze/Autoencoder.git

edited Feb 10 '16 at 12:28

serenesat

4,611
10
37
53

asked Feb 10 '16 at 12:10

Sn0w

Do you just want to extract the weights of the hidden units or see there activation? – Daniel Slater Feb 10 '16 at 13:40
I want to see there activation – Sn0w Feb 15 '16 at 11:11

how to extract (plot) the hidden units/softmax (features) from a autoencoder using Tensorflow

0 Answers0