Alert button

Modeling documents with Generative Adversarial Networks

Add code
Bookmark button
Alert button
Dec 29, 2016
John Glover

Figure 1 for Modeling documents with Generative Adversarial Networks
Figure 2 for Modeling documents with Generative Adversarial Networks
Figure 3 for Modeling documents with Generative Adversarial Networks
Figure 4 for Modeling documents with Generative Adversarial Networks

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

This paper describes a method for using Generative Adversarial Networks to learn distributed representations of natural language documents. We propose a model that is based on the recently proposed Energy-Based GAN, but instead uses a Denoising Autoencoder as the discriminator network. Document representations are extracted from the hidden layer of the discriminator and evaluated both quantitatively and qualitatively.

View paper onarxiv icon

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon