Models In a Spelling Bee: Language Models Implicitly Learn the Character Composition of Tokens

Anonymous

Paper Link: https://openreview.net/forum?id=5zxEabUyW9K

Paper Type: Short paper (up to four pages of content + unlimited references and appendices)

Abstract:

Standard pretrained language models operate
on sequences of subword tokens without direct access to the characters that compose each
token’s string representation. We probe the
embedding layer of pretrained language models and show that models learn the internal
character composition of whole word and subword tokens to a surprising extent, without
ever seeing the characters coupled with the tokens. Our results show that the embedding layers of RoBERTa and GPT2 each hold enough
information to accurately spell up to a third
of the vocabulary and reach high character
ngram overlap across all token types. We further test whether enriching subword models
with character information can improve language modeling, and observe that this method
has a near-identical learning curve as training without spelling-based enrichment. Overall, our results suggest that language modeling objectives incentivize the model to implicitly learn some notion of spelling, and that explicitly teaching the model how to spell does
not appear to enhance its performance on such
tasks.

Presentation Mode: This paper will be presented in person in Seattle

Virtual Presentation Timezone: UTC+2

Copyright Consent Signature (type Name Or NA If Not Transferrable): Itay Itzhak

Copyright Consent Name And Address: Tel Aviv University, Ramat Aviv 69978, Israel