Adjusting Interpretable Dimensions in Embedding Space with Human Judgments

Katrin Erk, Marianna Apidianaki


Abstract
Embedding spaces contain interpretable dimensions indicating gender, formality in style, or even object properties. This has been observed multiple times. Such interpretable dimensions are becoming valuable tools in different areas of study, from social science to neuroscience. The standard way to compute these dimensions uses contrasting seed words and computes difference vectors over them. This is simple but does not always work well. We combine seed-based vectors with guidance from human ratings of where words fall along a specific dimension, and evaluate on predicting both object properties like size and danger, and the stylistic properties of formality and complexity. We obtain interpretable dimensions with markedly better performance especially in cases where seed-based dimensions do not work well.
Anthology ID:
2024.naacl-long.146
Volume:
Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers)
Month:
June
Year:
2024
Address:
Mexico City, Mexico
Editors:
Kevin Duh, Helena Gomez, Steven Bethard
Venue:
NAACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2675–2686
Language:
URL:
https://aclanthology.org/2024.naacl-long.146
DOI:
Bibkey:
Cite (ACL):
Katrin Erk and Marianna Apidianaki. 2024. Adjusting Interpretable Dimensions in Embedding Space with Human Judgments. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers), pages 2675–2686, Mexico City, Mexico. Association for Computational Linguistics.
Cite (Informal):
Adjusting Interpretable Dimensions in Embedding Space with Human Judgments (Erk & Apidianaki, NAACL 2024)
Copy Citation:
PDF:
https://aclanthology.org/2024.naacl-long.146.pdf
Copyright:
 2024.naacl-long.146.copyright.pdf