@inproceedings{wen-etal-2026-benchmark,
    title = "The Benchmark Illusion: Pruned {LLM}s Can Pass Multiple Choice but Fail to Answer",
    author = "Wen, Rui  and
      Sun, Lu  and
      Liu, Jiayang  and
      Xu, Zesheng  and
      Cong, Tianshuo  and
      Li, Zheng",
    editor = "Lee, Hung-Yi  and
      Dev, Sunipa  and
      Li, Jessy  and
      Elliott, Desmond",
    booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2026",
    address = "Budapest, Hungary",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.emnlp-main.2313/",
    pages = "52565--52575",
    ISBN = "979-8-89176-459-0"
}