Publication: Why neural translations are the right length
Program
KU-Authors
KU Authors
Co-Authors
Shi, Xing
Knight, Kevin
Publication Date
Language
Embargo Status
Journal Title
Journal ISSN
Volume Title
Alternative Title
Abstract
We investigate how neural, encoder-decoder translation systems output target strings of appropriate lengths, finding that a collection of hidden units learns to explicitly implement this functionality.
Source
Publisher
Association for Computational Linguistics (ACL)
Subject
Computer engineering
Citation
Has Part
Source
EMNLP 2016 - Conference on Empirical Methods in Natural Language Processing, Proceedings
Book Series Title
Edition
DOI
10.18653/v1/d16-1248