@inproceedings{xu-etal-2025-obliviate,
    title = "{OBLIVIATE}: Robust and Practical Machine Unlearning for Large Language Models",
    author = "Xu, Xiaoyu  and
      Du, Minxin  and
      Ye, Qingqing  and
      Hu, Haibo",
    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.183/",
    pages = "3696--3715",
    ISBN = "979-8-89176-332-6",
    abstract = "Large language models (LLMs) trained over extensive corpora risk memorizing sensitive, copyrighted, or toxic content. To address this, we propose \textbf{OBLIVIATE}, a robust unlearning framework that removes targeted data while preserving model utility. The framework follows a structured process: extracting target tokens, building retain sets, and fine-tuning with a tailored loss function comprising three components{---}masking, distillation, and world fact. Using low-rank adapters (LoRA) ensures efficiency without compromising unlearning quality. We conduct experiments on multiple datasets, including Harry Potter series, WMDP, and TOFU, using a comprehensive suite of metrics: \textit{forget quality} (via a new document-level memorization score), \textit{model utility}, and \textit{fluency}. Results demonstrate its effectiveness in resisting membership inference attacks, minimizing the impact on retained data, and maintaining robustness across diverse scenarios."
}