Parallel Language Models via Variational Decoding [pdf]
Comments
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Summary
Parallel Language Models (PLMs) generate several mutually consistent tokens in one forward pass by routing dependencies through latent auxiliary variables, trained directly from data with a variational objective.