{"ID":5937714,"CreatedAt":"2026-07-07T03:14:33.014478982Z","UpdatedAt":"2026-07-08T14:31:46.746499803Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2607.04339","arxiv_id":"2607.04339","title":"One Framework for All: Cross-Modal Membership Inference for Generative Models","abstract":"Large generative models across text-to-text, text-to-image, and image-to-text modalities have been shown to pose significant privacy risks. One fundamental threat is membership inference attacks (MIA), which aim to determine whether a given data point was used in a model's training set. Although prior work has investigated MIAs against these three classes of generative models, existing approaches treat them in isolation and are not cross-applicable, thereby limiting their real-world utility. To address this limitation, we present the first comprehensive study of a unified membership inference framework that applies across text-to-text, text-to-image, and image-to-text modalities. Our approach is grounded in a key modality-agnostic observation: the output distribution of a generative model can approximate its training data distribution. Leveraging this property, we model the distributions of model-generated outputs and auxiliary non-member samples in a shared embedding space, and perform membership inference via likelihood ratio testing. We conduct extensive experiments in a strict black-box setting under both partial-knowledge and zero-knowledge threat models, and evaluate membership inference against both fine-tuning and pre-training data. Experimental results demonstrate our approach's superior performance in comparison to existing state-of-the-art methods, which are typically optimized for a single model class.","short_abstract":"Large generative models across text-to-text, text-to-image, and image-to-text modalities have been shown to pose significant privacy risks. One fundamental threat is membership inference attacks (MIA), which aim to determine whether a given data point was used in a model's training set. Although prior work has investig...","url_abs":"https://arxiv.org/abs/2607.04339","url_pdf":"https://arxiv.org/pdf/2607.04339v1","authors":"[\"Dayong Ye\",\"Tainqing Zhu\",\"Kun Gao\",\"Junhao Liu\",\"Yichuan Chen\",\"Shuai Zhou\",\"Hengzhu Liu\",\"Bo Liu\",\"Wanlei Zhou\"]","published":"2026-07-05T14:45:29Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.AI\",\"cs.CR\"]","methods":"[]","has_code":false}