Hackernews
new
show
ask
jobs
Parallel Language Models via Variational Decoding
3 points
posted 9 hours ago
by E-Reverance
(arxiv.org)
No comments yet