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.

翻译：大型语言模型（LLMs）已被证明会传播并放大有害的刻板印象，尤其是那些对边缘化群体影响尤为严重的偏见。为了更全面地理解这些刻板印象的影响，我们引入了GlobalBias数据集，该数据集包含87.6万个句子，涵盖了40个不同的性别-种族组合群体，并整合了偏见文献中常用的描述符，使我们能够研究来自世界各地的广泛刻板印象。我们利用GlobalBias通过困惑度直接探测一系列语言模型，将其作为衡量特定刻板印象在模型内部表征中如何体现的代理指标。在此基础上，我们基于给定姓名生成角色画像，并评估模型输出中刻板印象的普遍性。我们发现，与各类刻板印象相关的人口统计学群体在模型似然度和模型输出中均保持一致。此外，即使被明确指示避免偏见，规模更大的模型仍持续表现出更高程度的刻板化输出。