Unless I'm mistaken it seems like 
theano.tensor.sum(theano.tensor.exp(-10 * ws))
just encourages weights to become more positive.  Why not use a L2 weight 
penalty like
 theano.tensor.sum(w**2)
? 
So the full loss would become:
crossentropy_categorical_1hot(coding_dist, true_dist) + sum((layer.w**2).sum
() for layer in l_layers)

-- 

--- 
You received this message because you are subscribed to the Google Groups 
"theano-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
For more options, visit https://groups.google.com/d/optout.

Reply via email to