Quote from “Large Language Models Portray Socially Subordinate Groups as More Homogeneous, Consistent with a Bias Observed in Humans | Proceedings of the 2024 ACM Conference on Fairness, Accountability, and Transparency”