Modelling Implicit Bias in Gender–Career Associations: A systematic comparison of language models

Résumé fourni par la source

Biases in language and their reflection in language models have attracted researchers' attention, particularly with the growth of large language models (LLMs). However, many questions on the links between language models and people’s biased attitudes remain unanswered. In the current preregistered study, we focus on gender–career bias to examine the extent to which language models can be used to model behavioural responses in the Gender–Career Implicit Association Test (IAT). We provide a systematic evaluation of a range of language models, including n-gram, count vector, predict (word2vec), and Large Language Models (LLMs), to determine how well they capture people’s behaviour in the IAT. We compared response time data from over 800,000 participants from the UK and US against 25 language models, with a total of 675 model variants. We find that many language models, including large language models (LLMs), correlated well with human behavior. While results support previous findings for both predict and count model families, we observed that performance of LLMs was consistently different from that of simpler predict models, frequently exhibiting correlations in the opposite direction. Thus, our findings reinforce the idea that societal biases are generally encoded in language, but that large language models can exhibit behaviors different to classical language models. Study preregistration, code, and materials are all available from the project website.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.