Pith. sign in

REVIEW 7 cited by

Failures to Find Transferable Image Jailbreaks Between Vision-Language Models

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2407.15211 v2 pith:RDTP3Q2I submitted 2024-07-21 cs.CL cs.AIcs.CRcs.CVcs.LG

Failures to Find Transferable Image Jailbreaks Between Vision-Language Models

Rylan Schaeffer , Dan Valentine , Luke Bailey , James Chua , Crist\'obal Eyzaguirre , Zane Durante , Joe Benton , Brando Miranda
show 7 more authors
Henry Sleight John Hughes Rajashree Agrawal Mrinank Sharma Scott Emmons Sanmi Koyejo Ethan Perez
This is my paper
classification cs.CL cs.AIcs.CRcs.CVcs.LG
keywords vlmsimagejailbreakstransfermodelstransferablegradient-basedlanguage
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The integration of new modalities into frontier AI systems offers exciting capabilities, but also increases the possibility such systems can be adversarially manipulated in undesirable ways. In this work, we focus on a popular class of vision-language models (VLMs) that generate text outputs conditioned on visual and textual inputs. We conducted a large-scale empirical study to assess the transferability of gradient-based universal image ``jailbreaks" using a diverse set of over 40 open-parameter VLMs, including 18 new VLMs that we publicly release. Overall, we find that transferable gradient-based image jailbreaks are extremely difficult to obtain. When an image jailbreak is optimized against a single VLM or against an ensemble of VLMs, the jailbreak successfully jailbreaks the attacked VLM(s), but exhibits little-to-no transfer to any other VLMs; transfer is not affected by whether the attacked and target VLMs possess matching vision backbones or language models, whether the language model underwent instruction-following and/or safety-alignment training, or many other factors. Only two settings display partially successful transfer: between identically-pretrained and identically-initialized VLMs with slightly different VLM training data, and between different training checkpoints of a single VLM. Leveraging these results, we then demonstrate that transfer can be significantly improved against a specific target VLM by attacking larger ensembles of ``highly-similar" VLMs. These results stand in stark contrast to existing evidence of universal and transferable text jailbreaks against language models and transferable adversarial attacks against image classifiers, suggesting that VLMs may be more robust to gradient-based transfer attacks.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 7 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Jailbreaking Frontier Foundation Models Through Intention Deception

    cs.CR 2026-04 unverdicted novelty 7.0

    A multi-turn intention-deception jailbreak achieves high success on GPT-5 and Claude models while exposing para-jailbreaking where models leak harmful information without direct refusal.

  2. Visual Token Compression Enhances Robustness of MLLMs

    cs.CV 2026-07 conditional novelty 6.0

    Pruning visual tokens farthest from the text feature space at selected 'robust' layers improves MLLM jailbreak defense (average +13.29% RAR) and slightly reduces hallucination.

  3. DMN: A Compositional Framework for Jailbreaking Multimodal LLMs with Multi-Image Inputs

    cs.CR 2026-05 unverdicted novelty 6.0

    DMN achieves over 90% attack success rate on GPT-4o, Gemini-2.5-pro and Claude Sonnet 4 by distributing instructions, supplying multimodal evidence, and adding number chain tasks across multiple images.

  4. CaMeLs Can Use Computers Too: System-level Security for Computer Use Agents

    cs.AI 2026-01 conditional novelty 6.0

    Computer-use agents can run under Dual-LLM isolation with single-shot branching plans, preserving partial utility while blocking instruction injection, but remain open to branch-steering attacks.

  5. VISOR++: Universal Visual Inputs based Steering for Large Vision Language Models

    cs.CV 2025-09 conditional novelty 6.0

    A single adversarially optimized image can reproduce activation-steering behavior in multiple VLMs and partially transfer to unseen models.

  6. Laundering AI Authority with Adversarial Examples

    cs.CR 2026-05 unverdicted novelty 5.0

    Adversarial examples enable AI authority laundering by causing production VLMs to give authoritative but wrong responses on subtly perturbed images, with success rates of 22-100% using decade-old attack methods.

  7. ReVision : A Post-Hoc, Vision-Based Technique for Replacing Unacceptable Concepts in Image Generation Pipeline

    cs.CR 2026-02 conditional novelty 4.0

    ReVision uses a vision-language model's bounding box to gate attention-based image editing, suppressing unsafe concepts while better preserving benign background in multi-concept scenes.