@inproceedings{chen-etal-2025-safeeraser,
    title = "{S}afe{E}raser: Enhancing Safety in Multimodal Large Language Models through Multimodal Machine Unlearning",
    author = "Chen, Junkai  and
      Deng, Zhijie  and
      Zheng, Kening  and
      Yan, Yibo  and
      Liu, Shuliang  and
      Wu, PeiJun  and
      Jiang, Peijie  and
      Liu, Jia  and
      Hu, Xuming",
    editor = "Che, Wanxiang  and
      Nabende, Joyce  and
      Shutova, Ekaterina  and
      Pilehvar, Mohammad Taher",
    booktitle = "Findings of the Association for Computational Linguistics: ACL 2025",
    month = jul,
    year = "2025",
    address = "Vienna, Austria",
    publisher = "Association for Computational Linguistics",
    url = "https://preview.aclanthology.org/ingestion-acl-25/2025.findings-acl.731/",
    pages = "14194--14224",
    ISBN = "979-8-89176-256-5",
    abstract = "As Multimodal Large Language Models (MLLMs) develop, their potential security issues have become increasingly prominent. **Machine Unlearning (MU)**, as an effective strategy for forgetting specific knowledge in training data, has been widely used in privacy protection. However, *MU for safety in MLLM has yet to be fully explored*. To address this issue, we propose , a safety unlearning benchmark for MLLMs, consisting of 3,000 images and 28.8K VQA pairs. We comprehensively evaluate unlearning methods from two perspectives: **{\_}forget quality{\_}** and **{\_}model utility{\_}**. Our findings show that existing MU methods struggle to maintain model performance while implementing the forget operation and often suffer from **{\_}over-forgetting{\_}**. Hence, we introduce **Prompt Decouple (PD) Loss** to alleviate over-forgetting through decouple prompt during unlearning process. To quantitatively measure over-forgetting mitigated by PD Loss, we propose a new metric called **Safe Answer Refusal Rate (SARR)**. Experimental results demonstrate that combining PD Loss with existing unlearning methods can effectively prevent over-forgetting and achieve a decrease of 79.5{\%} in the SARR metric of LLaVA-7B and LLaVA-13B, while maintaining forget quality and model utility. Our code and dataset will be released upon acceptance. **Warning: This paper contains examples of harmful language and images, and reader discretion is recommended.**"
}