UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Stereotype and skew: Quantifying gender bias in pre-trained and fine-tuned language models

de Vassimon Manela, D; Errington, D; Fisher, T; van Breugel, B; Minervini, P; (2021) Stereotype and skew: Quantifying gender bias in pre-trained and fine-tuned language models. In: Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume. (pp. pp. 2232-2242). Association for Computational Linguistics Green open access

[thumbnail of 2021.eacl-main.190.pdf]
Preview
Text
2021.eacl-main.190.pdf - Published Version

Download (348kB) | Preview

Abstract

This paper proposes two intuitive metrics, skew and stereotype, that quantify and analyse the gender bias present in contextual language models when tackling the WinoBias pronoun resolution task. We find evidence that gender stereotype correlates approximately negatively with gender skew in out-of-the-box models, suggesting that there is a trade-off between these two forms of bias. We investigate two methods to mitigate bias. The first approach is an online method which is effective at removing skew at the expense of stereotype. The second, inspired by previous work on ELMo, involves the fine-tuning of BERT using an augmented gender-balanced dataset. We show that this reduces both skew and stereotype relative to its unaugmented fine-tuned counterpart. However, we find that existing gender bias benchmarks do not fully probe professional bias as pronoun resolution may be obfuscated by cross-correlations from other manifestations of gender prejudice. Our code is available online.

Type: Proceedings paper
Title: Stereotype and skew: Quantifying gender bias in pre-trained and fine-tuned language models
Event: EACL 2021 - 16th Conference of the European Chapter of the Association for Computational Linguistics
ISBN-13: 9781954085022
Open access status: An open access version is available from UCL Discovery
Publisher version: https://www.aclweb.org/anthology/2021.eacl-main.19...
Language: English
Additional information: ACL materials are Copyright © 1963–2021 ACL; other materials are copyrighted by their respective copyright holders. Materials prior to 2016 here are licensed under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 International License. Permission is granted to make copies for the purposes of teaching and research. Materials published in or after 2016 are licensed on a Creative Commons Attribution 4.0 International License.
UCL classification: UCL
UCL > Provost and Vice Provost Offices
UCL > Provost and Vice Provost Offices > UCL BEAMS
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science
UCL > Provost and Vice Provost Offices > UCL BEAMS > Faculty of Engineering Science > Dept of Computer Science
URI: https://discovery.ucl.ac.uk/id/eprint/10129955
Downloads since deposit
57Downloads
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item