Autoencoding
method
ID: autoencoding
Related Papers
Modeling Human Reading with Neural Attention
When humans read text, they fixate some words and skip others. However, there have been few attempts to explain skipping behav- ior with computational models, as most ex- isting work …...
Learning Word Meta-Embeddings by Autoencoding
AbstractDistributed word embeddings have shown superior performances in numerous Natural Language Processing (NLP) tasks. However, their performances vary significantly across different tasks, implying that the word embeddings learnt by those …...
Adversarial Learning on the Latent Space for Diverse Dialog Generation
AbstractGenerating relevant responses in a dialog is challenging, and requires not only proper modeling of context in the conversation, but also being able to generate fluent sentences during inference. In …...