Bai, X., Wang, A., Sucholutsky, I., Griffiths, T. L. (2025). Explicitly unbiased large language models still form biased associations. Proceedings of the National Academy of Sciences. https://doi.org/10.1073/pnas.2416228122