ReadTwice: Reading Very Large Documents with Memories
Yury Zemlyanskiy, Joshua Ainslie, Michiel de Jong, Philip Pham, Ilya Eckstein, Fei Sha
Abstract
Knowledge-intensive tasks such as question answering often require assimilating information from different sections of large inputs such as books or article collections. We propose ReadTwice, a simple and effective technique that combines several strengths of prior approaches to model long-range dependencies with Transformers. The main idea is to read text in small segments, in parallel, summarizing each segment into a memory table to be used in a second read of the text. We show that the method outperforms models of comparable size on several question answering (QA) datasets and sets a new state of the art on the challenging NarrativeQA task, with questions about entire books.
BibTeX
@inproceedings{zemlyanskiy-etal-2021-readtwice,
title = "{R}ead{T}wice: Reading Very Large Documents with Memories",
author = "Zemlyanskiy, Yury and
Ainslie, Joshua and
de Jong, Michiel and
Pham, Philip and
Eckstein, Ilya and
Sha, Fei",
editor = "Toutanova, Kristina and
Rumshisky, Anna and
Zettlemoyer, Luke and
Hakkani-Tur, Dilek and
Beltagy, Iz and
Bethard, Steven and
Cotterell, Ryan and
Chakraborty, Tanmoy and
Zhou, Yichao",
booktitle = "Proceedings of the 2021 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies",
month = jun,
year = "2021",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2021.naacl-main.408/",
doi = "10.18653/v1/2021.naacl-main.408",
pages = "5189--5195"
}