@inproceedings{oh-etal-2025-uncovering,
    title = "Uncovering Factor-Level Preference to Improve Human-Model Alignment",
    author = "Oh, Juhyun  and
      Kim, Eunsu  and
      Kim, Jiseon  and
      Xu, Wenda  and
      Cha, Inha  and
      Wang, William Yang  and
      Oh, Alice",
    editor = "Christodoulopoulos, Christos  and
      Chakraborty, Tanmoy  and
      Rose, Carolyn  and
      Peng, Violet",
    booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2025",
    month = nov,
    year = "2025",
    address = "Suzhou, China",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/author-page-yu-wang-polytechnic/2025.findings-emnlp.1045/",
    doi = "10.18653/v1/2025.findings-emnlp.1045",
    pages = "19179--19203",
    ISBN = "979-8-89176-335-7",
    abstract = "Large language models (LLMs) often exhibit tendencies that diverge from human preferences, such as favoring certain writing styles or producing overly verbose outputs. While crucial for improvement, identifying the factors driving these misalignments remains challenging due to existing evaluation methods' reliance on coarse-grained comparisons and lack of explainability.To address this, we introduce PROFILE, an automated framework to uncover and measure factor-level preference alignment of humans and LLMs.Using PROFILE, we analyze preference alignment across three key tasks: summarization, instruction-following, and document-based QA. We find a significant discrepancy: while LLMs show poor factor-level alignment with human preferences when generating texts, they demonstrate strong alignment in discrimination tasks. We demonstrate how leveraging the identified generation-discrimination gap can be used to improve LLM alignment through multiple approaches, including fine-tuning with self-guidance.Our work highlights the value of factor-level analysis for identifying hidden misalignments and provides a practical framework for improving LLM-human preference alignment."
}