Jimin Sun

Why are Sequence-to-Sequence Models So Dull?

Variational Autoregressive Decoder for Neural Response Generation

Content Preserving Text Generation with Attribute Controls

Matching Networks for One Shot Learning

Pointer Networks

next 