Lune

EMNLP Findings2024

Cross-modality Information Check for Detecting Jailbreaking in Multimodal Large Language Models

Yue Xu, Xiuyuan Qi, Zhan Qin, Wenjie Wang

Abstract

Multimodal Large Language Models (MLLMs) extend the capacity of LLMs to understand multimodal information comprehensively, achieving remarkable performance in many visioncentric tasks. Despite that, recent studies have shown that these models are susceptible to jailbreak attacks, which refer to an exploitative technique where malicious users can break the safety alignment of the target model and generate misleading and harmful answers. This potential threat is caused by both the inherent vulnerabilities of LLM and the larger attack scope introduced by vision input. To enhance the security of MLLMs against jailbreak attacks, researchers have developed various defense techniques. However, these methods either require modifications to the model's internal structure or demand significant computational resources during the inference phase. Multimodal information is a double-edged sword. While it increases the risk of attacks, it also provides additional data that can enhance safeguards. Inspired by this, we propose Cross-modality Information DEtectoR (CIDER), a plug-andplay jailbreaking detector designed to identify maliciously perturbed image inputs, utilizing the cross-modal similarity between harmful queries and adversarial images. CIDER is independent of the target MLLMs and requires less computation cost. Extensive experimental results demonstrate the effectiveness and efficiency of CIDER, as well as its transferability to both white-box and blackbox MLLMs. The resource is available at https://github.com/PandragonXIII/CIDER.

Ask Lune

Know this paper in minutes, not an afternoon.

Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.

16 top-tier papers have cited it since. Know what they found before your reviewers do.

Questions to start from

Lune can draw on

the full paper · the top-tier papers that cite it · related work

Get the answer

Free to start. No credit card required.

Lineage

  • Backdoor Cleaning without External Guidance in MLLM Fine-tuning cites MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models
  • InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning cites MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models
  • Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive Scoring cites Visual Instruction Tuning
  • SafeGRPO: Self-Rewarded Multimodal Safety Alignment via Rule-Governed Policy Optimization cites Backdoor Cleaning without External Guidance in MLLM Fine-tuning
  • Attention! Your Vision Language Model Could Be Maliciously Manipulated cites Visual Instruction Tuning
  • Jailbreaking Multimodal Large Language Models via Shuffle Inconsistency cites MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Models
Builds onCited byLeft to right by year; the more cited, the higher.

Cited by top-tier papers16

  1. Backdoor Cleaning without External Guidance in MLLM Fine-tuningXuankun Rong, Wenke Huang, Jian Liang, Jinhe Bi et al.NeurIPS 2025 · 39 citations
  2. SafeGRPO: Self-Rewarded Multimodal Safety Alignment via Rule-Governed Policy OptimizationXuankun Rong, Wenke Huang, Tingfeng Wang, Daiguo Zhou et al.CVPR 2026 · 13 citations
  3. Attention! Your Vision Language Model Could Be Maliciously ManipulatedXiaosen Wang, Shaokang Wang, Zhijin Ge, Yuyang Luo et al.NeurIPS 2025 · 13 citations
  4. Rethinking Jailbreak Detection of Large Vision Language Models with Representational Contrastive ScoringPeichun Hua, Hao Li, Shanghao Shi, Zhiyuan Yu et al.ACL 2026 · 8 citations
  5. Jailbreaking Multimodal Large Language Models via Shuffle InconsistencyShiji Zhao, Ranjie Duan, Fengxiang Wang, Chi Chen et al.ICCV 2025 · 8 citations

Builds on14

  1. An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
  2. Visual Instruction TuningHaotian Liu, Chunyuan Li, Qingyang Wu, Yong Jae LeeNeurIPS 2023 · 11,349 citations
  3. Improved Denoising Diffusion Probabilistic ModelsAlexander Quinn Nichol, Prafulla DhariwalICML 2021 · 5,234 citations
  4. InstructBLIP: Towards General-purpose Vision-Language Models with Instruction TuningWenliang Dai, Junnan Li, Dongxu Li, Anthony Meng Huat Tiong et al.NeurIPS 2023 · 4,013 citations
  5. MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language ModelsDeyao Zhu, Jun Chen, Xiaoqian Shen, Xiang Li et al.ICLR 2024 · 3,079 citations

Related papers

  1. Efficient LLM-Jailbreaking via Multimodal-LLM JailbreakHaoxuan Ji, Zheng Lin, Zhenxing Niu, Xinbo Gao et al.AAAI 2026 · 4 citations
  2. MIDAS: Multi-Image Dispersion and Semantic Reconstruction for Jailbreaking MLLMsYilian Liu, Guoshun Nan, Jiuyang Lyu, Zhican Chen et al.ICLR 2026 · 3 citations
  3. Heuristic-Induced Multimodal Risk Distribution Jailbreak Attack for Multimodal Large Language ModelsTeng Ma, Xiaojun Jia, Ranjie Duan, Xinfeng Li et al.ICCV 2025 · 35 citations
  4. Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language ModelsErfan Shayegani, Yue Dong, Nael B. Abu-GhazalehICLR 2024 · 271 citations
  5. Odysseus: Jailbreaking Commercial Multimodal LLM-integrated Systems via Dual SteganographySongze Li, Jiameng Cheng, Yiming Li, Xiaojun Jia et al.NDSS 2026 · 9 citations
Dusk over the sea between two cliffs drawn in fine vertical lines