Think Again Networks and the Delta Loss

Add code
Apr 30, 2019
Alexandre Salle, Marcelo Prates
Figure 1 for Think Again Networks and the Delta Loss
Figure 2 for Think Again Networks and the Delta Loss

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

This short paper introduces an abstraction called Think Again Networks (ThinkNet) which can be applied to any state-dependent function (such as a recurrent neural network).

* redacted experiments on language modeling due to evaluation error  
View paper onarxiv icon

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon