Metraipolis › archive › dc_c80cb2e6b58e

Understanding In-Context Multimodal Jailbreaks via Posterior Reweighting — arXiv:2609.10613

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.10613
License (as archived)
CC0-1.0
Archived
2026-10-16T05:46:20.246Z
SHA-256 of the archived text
41d39106f4c095d03476e51882892059712e2cad72b30ef782f4ab9bc4e485a7
Stable URL of this page
https://metraipolis.com/documento/dc_c80cb2e6b58e

How to cite

Understanding In-Context Multimodal Jailbreaks via Posterior Reweighting — arXiv:2609.10613 — arXiv (en), https://arxiv.org/abs/2609.10613 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_c80cb2e6b58e (Metraipolis archive dc_c80cb2e6b58e, sha256 41d39106f4c095d03476e51882892059712e2cad72b30ef782f4ab9bc4e485a7)
BibTeX
@misc{dcc80cb2e6b58e,
  title = {Understanding In-Context Multimodal Jailbreaks via Posterior Reweighting — arXiv:2609.10613},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.10613},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_c80cb2e6b58e; sha256 41d39106f4c095d03476e51882892059712e2cad72b30ef782f4ab9bc4e485a7},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.10613. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In-context learning (ICL) jailbreaks reveal a critical vulnerability in multimodal large language models (MLLMs): harmful demonstrations in the prompt can induce unsafe outputs without modifying model parameters. Despite extensive empirical evidence, existing work lacks a principled understanding of why such jailbreaks reliably succeed or how their effectiveness scales with context composition. We propose a posterior reweighting framework that models a safety-aligned MLLM as implicitly operating over competing behavioral modes, and interprets in-context demonstrations as inference-time evidence that dynamically shifts the model's posterior preference between safe and harmful behaviors. This view formalizes jailbreak as a process of evidence accumulation, yielding predictive scaling laws with respect to demonstration count, harmful ratio, adversarial strength, and semantic diversity. Guided by this framework, we introduce a posterior-aware inference-time defense that adaptively injects benign counter-evidence based on estimated risk, effectively suppressing harmful posterior drift while preserving model utility. Compared to existing in-context defenses, our method achieves a significantly improved robustness-utility trade-off under a fixed intervention budget. Together, our results establish posterior reweighting as a unifying and predictive framework for understanding and mitigating ICL jailbreak in MLLMs.

Tema degli esploratori: fiducia digitale e crittografia (retro-catalogo arXiv: annunci del 2026-09-11 in cs.CR).
Categorie arXiv: Cryptography and Security (cs.CR) ; Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2609.10613 — 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 2026-10-16T05:30:39.369Z dalla pagina https://arxiv.org/catchup/cs.CR/2026-09-11?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).