Women Are Beautiful, Men Are Leaders: Gender Stereotypes in Machine Translation and Language Modeling

TL;DR

This paper introduces GEST, a dataset for measuring gender stereotypes in language models and machine translation, revealing that larger models tend to exhibit more stereotypical reasoning across multiple languages.

Contribution

The creation of GEST dataset and its application to evaluate gender stereotypes in various language models and translation systems, highlighting the prevalence of stereotypes.

Findings

Models show significant gender-stereotypical reasoning.

Larger models tend to be more stereotypical.

Stereotypes are consistent across languages.

Abstract

We present GEST -- a new manually created dataset designed to measure gender-stereotypical reasoning in language models and machine translation systems. GEST contains samples for 16 gender stereotypes about men and women (e.g., Women are beautiful, Men are leaders) that are compatible with the English language and 9 Slavic languages. The definition of said stereotypes was informed by gender experts. We used GEST to evaluate English and Slavic masked LMs, English generative LMs, and machine translation systems. We discovered significant and consistent amounts of gender-stereotypical reasoning in almost all the evaluated models and languages. Our experiments confirm the previously postulated hypothesis that the larger the model, the more stereotypical it usually is.