A temperature-perturbed black-box attack infers video training membership in VideoLLMs with 0.68 AUC by exploiting sharper generation behavior on member samples.
arXiv preprint arXiv:2601.11210 (2026)
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Life-logging video streams create an inevitable privacy-utility trade-off that is a foundational challenge for always-on AI systems.
citing papers explorer
-
Membership Inference Attacks Against Video Large Language Models
A temperature-perturbed black-box attack infers video training membership in VideoLLMs with 0.68 AUC by exploiting sharper generation behavior on member samples.
-
Position: Life-Logging Video Streams Make the Privacy-Utility Trade-off Inevitable
Life-logging video streams create an inevitable privacy-utility trade-off that is a foundational challenge for always-on AI systems.