Who is better at math, Jenny or Jingzhen? Uncovering Stereotypes in Large Language Models

TL;DR

This paper introduces GlobalBias, a large dataset to analyze stereotypes in language models, revealing that bigger models tend to produce more stereotypical outputs across diverse demographic groups.

Contribution

The paper presents GlobalBias, a comprehensive dataset for studying stereotypes in LLMs, and systematically evaluates how model size influences stereotype propagation.

Findings

Larger models exhibit higher levels of stereotypical outputs.

Stereotypes remain consistent across model likelihoods and outputs.

GlobalBias enables broad analysis of stereotypes across 40 demographic groups.

Abstract

Large language models (LLMs) have been shown to propagate and amplify harmful stereotypes, particularly those that disproportionately affect marginalised communities. To understand the effect of these stereotypes more comprehensively, we introduce GlobalBias, a dataset of 876k sentences incorporating 40 distinct gender-by-ethnicity groups alongside descriptors typically used in bias literature, which enables us to study a broad set of stereotypes from around the world. We use GlobalBias to directly probe a suite of LMs via perplexity, which we use as a proxy to determine how certain stereotypes are represented in the model's internal representations. Following this, we generate character profiles based on given names and evaluate the prevalence of stereotypes in model outputs. We find that the demographic groups associated with various stereotypes remain consistent across model likelihoods and model outputs. Furthermore, larger models consistently display higher levels of stereotypical outputs, even when explicitly instructed not to.