Unsupervised Token-wise Alignment to Improve Interpretation of Encoder-Decoder models

Publication Information

Title

Japanese:
English:	Unsupervised Token-wise Alignment to Improve Interpretation of Encoder-Decoder models

Author

Japanese:	Shun Kiyono, Sho Takase, Jun Suzuki, 岡崎直観, Kentaro Inui, Masaaki Nagata.
English:	Shun Kiyono, Sho Takase, Jun Suzuki, Naoaki Okazaki, Kentaro Inui, Masaaki Nagata.

Language

English

Journal/Book name

Japanese:
English:

Volume, Number, Page

pp. 74-81

Published date

Nov. 2018

Publisher

Japanese:
English:

Conference name

Japanese:
English:	Proceedings of the Workshop on Analyzing and interpreting neural networks for NLP (BlackboxNLP)

Conference site

Japanese:
English: