Social-Cognitive Bias & Machine Learning
Traditional measures of the content of social groups’ stereotypes differ from measures of stereotype salience (Nicolas, Bai, & Fiske, in prep.)
We study how AI language and computer vision models reflect human biases. Developing projects examines stereotype content (both for single and intersectional identities) in archival data (books, news articles, social media). In addition, we use state-of-the-art face recognition models to generate realistic face photographs to study human biases in face impressions.
