@inproceedings{sorensen-etal-2025-value,
    title = "Value Profiles for Encoding Human Variation",
    author = "Sorensen, Taylor  and
      Mishra, Pushkar  and
      Patel, Roma  and
      Tessler, Michael Henry  and
      Bakker, Michiel A.  and
      Evans, Georgina  and
      Gabriel, Iason  and
      Goodman, Noah  and
      Rieser, Verena",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingest-emnlp/2025.emnlp-main.106/",
    pages = "2047--2095",
    ISBN = "979-8-89176-332-6",
    abstract = "Modelling human variation in rating tasks is crucial for enabling AI systems for personalization, pluralistic model alignment, and computational social science. We propose representing individuals using \textit{value profiles} {--} natural language descriptions of underlying values compressed from in-context demonstrations {--} along with a steerable decoder model to estimate ratings conditioned on a value profile or other rater information. To measure the predictive information in rater representations, we introduce an information-theoretic methodology. We find that demonstrations contain the most information, followed by value profiles and then demographics. However, value profiles offer advantages in terms of scrutability, interpretability, and steerability due to their compressed natural language format. Value profiles effectively compress the useful information from demonstrations (70{\%} information preservation). Furthermore, clustering value profiles to identify similarly behaving individuals better explains rater variation than the most predictive demographic groupings. Going beyond test set performance, we show that the decoder models interpretably change ratings according to semantic profile differences, are well-calibrated, and can help explain instance-level disagreement by simulating an annotator population. These results demonstrate that value profiles offer novel, predictive ways to describe individual variation beyond demographics or group information."
}