Hacker News new | past | comments | ask | show | jobs | submit login
Discrete Variational Autoencoders (arxiv.org)
24 points by rch on Sept 11, 2016 | hide | past | favorite | 1 comment



Trying to wrap my mind around the paper. Could be a really important finding. Should now allow for efficient use of discrete variables in deep neural network models.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: