@inproceedings{laskar-etal-2025-judging,
    title = "Judging the Judges: Can Large Vision-Language Models Fairly Evaluate Chart Comprehension and Reasoning?",
    author = "Laskar, Md Tahmid Rahman  and
      Islam, Mohammed Saidul  and
      Mahbub, Ridwan  and
      Masry, Ahmed  and
      Rahman, Mizanur  and
      Bhuiyan, Amran  and
      Nayeem, Mir Tafseer  and
      Joty, Shafiq  and
      Hoque, Enamul  and
      Huang, Jimmy",
    editor = "Rehm, Georg  and
      Li, Yunyao",
    booktitle = "Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 6: Industry Track)",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/landing_page/2025.acl-industry.83/",
    pages = "1203--1216",
    ISBN = "979-8-89176-288-6",
    abstract = "Charts are ubiquitous as they help people understand and reason with data. Recently, various downstream tasks, such as chart question answering, chart2text, and fact-checking, have emerged. Large Vision-Language Models (LVLMs) show promise in tackling these tasks, but their evaluation is costly and time-consuming, limiting real-world deployment. While using LVLMs as judges to assess chart comprehension capabilities of other LVLMs could streamline evaluation processes, challenges like proprietary datasets, restricted access to powerful models, and evaluation costs hinder their adoption in industrial settings. To this end, we present a comprehensive evaluation of 13 open-source LVLMs as judges for diverse chart comprehension and reasoning tasks. We design both pairwise and pointwise evaluation tasks covering criteria like factual correctness, informativeness, and relevancy. Additionally, we analyze LVLM judges based on format adherence, positional consistency, length bias, and instruction-following. We focus on cost-effective LVLMs ({\ensuremath{<}}10B parameters) suitable for both research and commercial use, following a standardized evaluation protocol and rubric to measure the LVLM judge accuracy. Experimental results reveal notable variability: while some open LVLM judges achieve GPT-4-level evaluation performance (about 80{\%} agreement with GPT-4 judgments), others struggle (below {\textasciitilde}10{\%} agreement). Our findings highlight that state-of-the-art open-source LVLMs can serve as cost-effective automatic evaluators for chart-related tasks, though biases such as positional preference and length bias persist."
}