Paper Detail

Open-MMUnlearning: Unifying Methods and Evaluation for MLLM Unlearning

Junkai Chen, Yuhao He, Qianshan Wei, Junxiang You, Jingwen Shao, Junkai Lin, Zhongkai Yue, Xiaotian Ye, Zhengbo Jiao, Jiali Cheng, Zhijie Deng, Kening Zheng, Ruiqi Liu, Hadi Amiri, Yi Yu, Zhenan Sun, Qi Li, Ka-Ho Chow, Sijia Liu, Liang Wang, Jiaqi Li, Shu Wu

arxiv Score 13.8

Published 2026-10-07 · First seen 2026-10-08

Research Track A · General AI

Abstract

As multimodal large language models (MLLMs) become more capable and widely deployed, concerns about privacy and safety have become increasingly pressing. Machine unlearning offers one approach to addressing these concerns by removing designated information from trained models while preserving unrelated capabilities. However, fragmented implementations and evaluation protocols, incomplete robustness testing, and limited understanding of metric reliability make progress in MLLM unlearning difficult to assess systematically. We introduce Open-MMUnlearning, an open-source, extensible framework that integrates target-model preparation, multimodal data processing, unlearning, and evaluation through shared interfaces and structured configurations. The framework supports five benchmarks spanning privacy, safety, and copyright, eight MLLMs from four model families, and twelve unlearning methods. Its evaluation suite jointly assesses forgetting effectiveness, retained utility, and robustness to model interventions, adversarial inputs, and membership inference attacks. Using a common evaluation protocol, we compare ten representative unlearning methods. In this comparison, GD and MIP-Editor tie for the highest overall score: GD achieves the highest Forget Quality, while MIP-Editor preserves more Model Utility. We further introduce a metric meta-evaluation protocol that tests faithfulness using models with controlled exposure to target knowledge and robustness under quantization and relearning. Among the thirteen evaluated metrics, BLEU achieves the highest aggregate reliability score. KS-Test attains the highest faithfulness AUC but performs less well on robustness. Together, the framework and these findings support reproducible comparison of MLLM unlearning methods and systematic assessment of evaluation reliability.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
now
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@article{chen2026open,
  title = {Open-MMUnlearning: Unifying Methods and Evaluation for MLLM Unlearning},
  author = {Junkai Chen and Yuhao He and Qianshan Wei and Junxiang You and Jingwen Shao and Junkai Lin and Zhongkai Yue and Xiaotian Ye and Zhengbo Jiao and Jiali Cheng and Zhijie Deng and Kening Zheng and Ruiqi Liu and Hadi Amiri and Yi Yu and Zhenan Sun and Qi Li and Ka-Ho Chow and Sijia Liu and Liang Wang and Jiaqi Li and Shu Wu},
  year = {2026},
  abstract = {As multimodal large language models (MLLMs) become more capable and widely deployed, concerns about privacy and safety have become increasingly pressing. Machine unlearning offers one approach to addressing these concerns by removing designated information from trained models while preserving unrelated capabilities. However, fragmented implementations and evaluation protocols, incomplete robustness testing, and limited understanding of metric reliability make progress in MLLM unlearning difficul},
  url = {https://arxiv.org/abs/2610.10358},
  keywords = {cs.AI},
  eprint = {2610.10358},
  archiveprefix = {arXiv},
}

Metadata

{}