Emergent linguistic structure in artificial neural networks trained by self-supervision

Christopher D. Manning*, Kevin Clark, John Hewitt, Urvashi Khandelwal, Omer Levy

*Corresponding author for this work

Research output: Contribution to journalArticlepeer-review

Abstract

This paper explores the knowledge of linguistic structure learned by large artificial neural networks, trained via self-supervision, whereby the model simply tries to predict a masked word in a given context. Human language communication is via sequences of words, but language understanding requires constructing rich hierarchical structures that are never observed explicitly. The mechanisms for this have been a prime mystery of human language acquisition, while engineering work has mainly proceeded by supervised learning on treebanks of sentences hand labeled for this latent structure. However, we demonstrate that modern deep contextual language models learn major aspects of this structure, without any explicit supervision. We develop methods for identifying linguistic hierarchical structure emergent in artificial neural networks and demonstrate that components in these models focus on syntactic grammatical relationships and anaphoric coreference. Indeed, we show that a linear transformation of learned embeddings in these models captures parse tree distances to a surprising degree, allowing approximate reconstruction of the sentence tree structures normally assumed by linguists. These results help explain why these models have brought such large improvements across many language-understanding tasks.

Original languageEnglish
Pages (from-to)30046-30054
Number of pages9
JournalProceedings of the National Academy of Sciences of the United States of America
Volume117
Issue number48
DOIs
StatePublished - 1 Dec 2020
Externally publishedYes

Funding

FundersFunder number
Google

    Keywords

    • Artificial neural netwok
    • Self-supervision
    • Syntax | learning

    Fingerprint

    Dive into the research topics of 'Emergent linguistic structure in artificial neural networks trained by self-supervision'. Together they form a unique fingerprint.

    Cite this