Metraipolis › archive › dc_04e0bbbbc3f5

TempJail: Temporal Jailbreak Attack against Large Vision-Language Models via Subtitle Scheduling — arXiv:2608.19737

Source
arXiv (en)
Original
https://arxiv.org/abs/2608.19737
License (as archived)
CC0-1.0
Archived
2026-10-18T11:39:29.837Z
SHA-256 of the archived text
20aa57fd387721ccd2be26b4691ecb6082c3cd0f2bfea06f06e86546d81e7259
Stable URL of this page
https://metraipolis.com/documento/dc_04e0bbbbc3f5

How to cite

TempJail: Temporal Jailbreak Attack against Large Vision-Language Models via Subtitle Scheduling — arXiv:2608.19737 — arXiv (en), https://arxiv.org/abs/2608.19737 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_04e0bbbbc3f5 (Metraipolis archive dc_04e0bbbbc3f5, sha256 20aa57fd387721ccd2be26b4691ecb6082c3cd0f2bfea06f06e86546d81e7259)
BibTeX
@misc{dc04e0bbbbc3f5,
  title = {TempJail: Temporal Jailbreak Attack against Large Vision-Language Models via Subtitle Scheduling — arXiv:2608.19737},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.19737},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_04e0bbbbc3f5; sha256 20aa57fd387721ccd2be26b4691ecb6082c3cd0f2bfea06f06e86546d81e7259},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.19737. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large vision-language models (LVLMs) have achieved remarkable progress in video understanding and reasoning. Despite extensive studies on text- and image-based jailbreaks, video jailbreaks against LVLMs remain largely unexplored. Existing video jailbreak methods mainly manipulate textual content embedded in videos, while overlooking how such information is organized over time. Our analysis reveals that jailbreak effectiveness depends not only on the semantics of textual information but also on its temporal presentation, including duration and timing-slot allocation. Motivated by this finding, we use subtitles, which are common in real-world videos and allow semantic content to be presented under precise temporal control without appearing visually intrusive, as a natural attack medium. Based on this insight, we propose TempJail, a black-box video-based jailbreak framework that constructs query-aligned dialogue-style subtitle sequences and optimizes their temporal scheduling to exploit temporal vulnerabilities in LVLMs and elicit responses that satisfy the harmful intent of the source query. Extensive experiments on four representative LVLMs and two datasets demonstrate that TempJail achieves the highest attack success rate across all evaluated model--dataset settings, outperforming the strongest baseline by 53 and 18 percentage points in dataset-averaged ASR on GPT-5 and Gemini 3.5-Flash, respectively.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-21 in cs.CL).
Categorie arXiv: Computer Vision and Pattern Recognition (cs.CV) ; Artificial Intelligence (cs.AI); Computation and Language (cs.CL).
Pagina del lavoro: https://arxiv.org/abs/2608.19737 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-18T11:38:27.948Z dalla pagina https://arxiv.org/catchup/cs.CL/2026-08-21?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).