Predict the Next Word: <Humans exhibit uncertainty in this task and language models _____>

Evgenia Ilia, Wilker Aziz


Abstract
Language models (LMs) are statistical models trained to assign probability to human-generated text. As such, it is reasonable to question whether they approximate linguistic variability exhibited by humans well. This form of statistical assessment is difficult to perform at the passage level, for it requires acceptability judgments (i.e., human evaluation) or a robust automated proxy (which is non-trivial). At the word level, however, given some context, samples from an LM can be assessed via exact matching against a prerecorded dataset of alternative single-word continuations of the available context. We exploit this fact and evaluate the LM’s ability to reproduce variability that humans (in particular, a population of English speakers) exhibit in the ‘next word prediction’ task. This can be seen as assessing a form of calibration, which, in the context of text classification, Baan et al. (2022) termed calibration to human uncertainty. We assess GPT2, BLOOM and ChatGPT and find that they exhibit fairly low calibration to human uncertainty. We also verify the failure of expected calibration error (ECE) to reflect this, and as such, advise the community against relying on it in this setting.
Anthology ID:
2024.eacl-short.22
Original:
2024.eacl-short.22v1
Version 2:
2024.eacl-short.22v2
Volume:
Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 2: Short Papers)
Month:
March
Year:
2024
Address:
St. Julian’s, Malta
Editors:
Yvette Graham, Matthew Purver
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
234–255
Language:
URL:
https://aclanthology.org/2024.eacl-short.22
DOI:
Bibkey:
Cite (ACL):
Evgenia Ilia and Wilker Aziz. 2024. Predict the Next Word: . In Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics (Volume 2: Short Papers), pages 234–255, St. Julian’s, Malta. Association for Computational Linguistics.
Cite (Informal):
Predict the Next Word: (Ilia & Aziz, EACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.eacl-short.22.pdf
Video:
 https://aclanthology.org/2024.eacl-short.22.mp4