Logit-based MIAs perform comparably on MLLMs with or without visual inputs in-distribution but visual inputs mask membership signals in out-of-distribution settings.
Membership inference attacks against large vision-language models.Advances in Neural Information Processing Systems, 37:98645–98674, 2024
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CR 1years
2025 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Lost in Modality: Evaluating the Effectiveness of Text-Based Membership Inference Attacks on Large Multimodal Models
Logit-based MIAs perform comparably on MLLMs with or without visual inputs in-distribution but visual inputs mask membership signals in out-of-distribution settings.