Pith. sign in

REVIEW 24 cited by

Jailbreaking Attack against Multimodal Large Language Model

Not yet reviewed by Pith; the record is open.

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 2402.02309 v1 pith:5WG3ILU7 submitted 2024-02-04 cs.LG cs.CLcs.CRcs.CV

Jailbreaking Attack against Multimodal Large Language Model

classification cs.LG cs.CLcs.CRcs.CV
keywords jailbreakinglanguagemllmsmodelsapproachgeneratedimgjplarge
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

This paper focuses on jailbreaking attacks against multi-modal large language models (MLLMs), seeking to elicit MLLMs to generate objectionable responses to harmful user queries. A maximum likelihood-based algorithm is proposed to find an \emph{image Jailbreaking Prompt} (imgJP), enabling jailbreaks against MLLMs across multiple unseen prompts and images (i.e., data-universal property). Our approach exhibits strong model-transferability, as the generated imgJP can be transferred to jailbreak various models, including MiniGPT-v2, LLaVA, InstructBLIP, and mPLUG-Owl2, in a black-box manner. Moreover, we reveal a connection between MLLM-jailbreaks and LLM-jailbreaks. As a result, we introduce a construction-based method to harness our approach for LLM-jailbreaks, demonstrating greater efficiency than current state-of-the-art methods. The code is available here. \textbf{Warning: some content generated by language models may be offensive to some readers.}

discussion (0)

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

Forward citations

Cited by 24 Pith papers

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

  1. Same Model, Different Weakness: How Language and Modality Reshape the Jailbreak Attack Surface in Frontier MLLMs

    cs.CL 2026-05 unverdicted novelty 8.0

    Jailbreak vulnerability in MLLMs is language- and modality-dependent, producing rank reversals in model safety between English and Spanish conditions.

  2. RouteHijack: Routing-Aware Attack on Mixture-of-Experts LLMs

    cs.LG 2026-05 unverdicted novelty 7.0

    RouteHijack is a routing-aware jailbreak that identifies safety-critical experts via activation contrast and optimizes suffixes to suppress them, reaching 69.3% average attack success rate on seven MoE LLMs with stron...

  3. MASCing: Configurable Mixture-of-Experts Behavior via Activation Steering Masks

    cs.CR 2026-04 unverdicted novelty 7.0

    MASCing uses an LSTM surrogate and optimized steering masks to enable flexible, inference-time control over MoE expert routing for safety objectives, improving jailbreak defense and content generation success rates su...

  4. DeepBias: Adaptive In-depth Probing of Social Biases in LVLMs

    cs.CY 2026-07 conditional novelty 6.5

    A generation-evolution-probing loop with Proposer and Digger agents produces DeepBiasBench, exposing deeper LVLM social biases than static image-question benchmarks.

  5. 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.

  6. 3D FaceShell: Attribute Transfer in 3D Face Avatars as a VLM Defense Mechanism

    cs.CV 2026-07 conditional novelty 6.0

    An optimized, view-consistent 3D Gaussian shell around a face avatar reliably changes which facial attributes vision-language models report, while preserving identity and appearance.

  7. PPE-Bench: A Benchmark for Evaluating MLLM Unlearning under Private-Public Entanglement

    cs.CR 2026-07 conditional novelty 6.0

    Existing MLLM unlearning methods reduce private-attribute leakage on entangled images but substantially harm co-occurring public figures and landmarks, with private knowledge often re-emerging after public finetuning.

  8. Safe responses matter: Output-aware safety guardrail mitigate over-refusal in MLLMs

    cs.LG 2026-06 conditional novelty 6.0

    A lightweight multi-instance contrastive classifier on MLLM hidden states predicts output harmfulness, matching input-side safety while sharply lowering over-refusal.

  9. Unveiling Privacy Risks in Multi-modal Large Language Models: Task-specific Vulnerabilities and Mitigation Challenges

    cs.CR 2026-06 unverdicted novelty 6.0

    Introduces MM-Privacy dataset and evaluations showing MLLMs leak sensitive data from images in various tasks, highlighting task inconsistency effects.

  10. Investigating Adversarial Robustness of Multi-modal Large Language Models

    cs.CV 2026-06 unverdicted novelty 6.0

    Robust vision encoders from multimodal adversarial pretraining transfer to MLLMs and deliver large gains in adversarial captioning and VQA performance, while test-time stochastic transformations provide an effective b...

  11. Attention Hijacking: Response Manipulation Across Queries in Vision-Language Models

    cs.CV 2026-05 unverdicted novelty 6.0

    Attention Hijacking is a new attack that improves cross-query transferability in VLMs by explicitly steering internal attention to a persistent image-dominant pattern.

  12. SafeSteer: A Decoding-level Defense Mechanism for Multimodal Large Language Models

    cs.AI 2026-05 unverdicted novelty 6.0

    SafeSteer improves safety in multimodal large language models by up to 33.4% via a decoding probe and modal alignment vector without any fine-tuning.

  13. Securing Multimodal AI through Internal Information Decomposition

    cs.AI 2026-05 conditional novelty 6.0

    A one-class detector using first-token distributional consistency between text-only, vision-only, and joint predictions reduces multimodal jailbreak attack success rates to below 15% with ~2.4% utility loss.

  14. Every Picture Tells a Dangerous Story: Memory-Augmented Multi-Agent Jailbreak Attacks on VLMs

    cs.AI 2026-04 unverdicted novelty 6.0

    MemJack achieves 71.48% attack success rate on unmodified COCO val2017 images against Qwen3-VL-Plus by coordinating agents to map visual entities to malicious intents, apply multi-angle camouflage, and filter refusals...

  15. TEMPLATEFUZZ: Fine-Grained Chat Template Fuzzing for Jailbreaking and Red Teaming LLMs

    cs.CR 2026-04 unverdicted novelty 6.0

    TEMPLATEFUZZ mutates chat templates with element-level rules and heuristic search to reach 98.2% average jailbreak success rate on twelve open-source LLMs while degrading accuracy by only 1.1%.

  16. RedDiffuser: Auditing Multimodal Safety Failures in Vision-Language Models via Reinforced Diffusion

    cs.CV 2025-03 unverdicted novelty 6.0

    RedDiffuser is a reinforced diffusion framework that generates adversarial visual contexts to audit and expose widespread multimodal safety failures in VLMs, increasing unsafe response rates by up to 10.69% on LLaVA w...

  17. PHANTOM: A Large-Scale Dataset of Multimodal Adversarial Attacks for Vision-Language Models

    cs.AI 2026-06 unverdicted novelty 5.0

    PHANTOM is a consolidated open-source dataset of 47,524 multimodal adversarial samples for VLMs, extending prior benchmarks across 10 high-level categories and 55 subcategories of harmful intents.

  18. Auditing Inference-Time Defense Evaluation for Multimodal Large Language Models

    cs.CR 2026-06 conditional novelty 5.0

    Stacking inference-time defenses for multimodal LLMs drives 97–100% over-refusal of benign queries; no single defense dominates and a simple prompt is often preferable.

  19. Auditing Inference-Time Defense Evaluation for Multimodal Large Language Models

    cs.CR 2026-06 conditional novelty 5.0

    An audit of a multimodal-LLM defense-evaluation archive finds three benchmark branches fail provenance checks, the surviving keyword-protocol results are not validated harmlessness rates, and the archive does not supp...

  20. SoK: Robustness in Large Language Models against Jailbreak Attacks

    cs.CR 2026-05 accept novelty 5.0

    The paper taxonomizes jailbreak attacks and defenses for LLMs, introduces the Security Cube multi-dimensional evaluation framework, benchmarks 13 attacks and 5 defenses, and identifies open challenges in LLM robustness.

  21. Stage-wise Attention-Guided Region Sequencing for Adversarial Attacks on Large Vision-Language Models

    cs.CV 2026-02 conditional novelty 5.0

    Attention-guided stage-wise cropping attacks beat random-crop baselines on targeted adversarial attacks for LVLMs, with less visible perturbation.

  22. Auditing Inference-Time Defense Evaluation for Multimodal Large Language Models

    cs.CR 2026-06 unverdicted novelty 4.0

    Empirical comparison finds no single inference-time defense dominates for MLLMs, combinations cause 97-100% over-refusal on benign queries, and adaptive selection based on model and attack type is recommended.

  23. Image-to-Video Diffusion: From Foundations to Open Frontiers

    cs.CV 2026-05 unverdicted novelty 3.0

    A survey that organizes diffusion image-to-video methods into a taxonomy, distills core designs in condition encoding, temporal modeling, noise prior, and upsampling, and discusses applications plus challenges.

  24. Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety

    cs.CR 2025-02 unverdicted novelty 2.0

    A comprehensive survey that taxonomizes safety threats to large models and agents, reviews defenses and benchmarks, and outlines open challenges.