mlsanityData to 7 Oct 2026

LAMP: Learning Universal Adversarial Perturbations for Multi-Image Tasks via Pre-trained Models

arXiv 2601.21220 ↗First posted 29 Jan 2026cs.CVcited by 0

On the map: Multimodal LLMs / vision-language models → MLLM jailbreaks and adversarial robustness

Abstract (arXiv)

Multimodal Large Language Models (MLLMs) have achieved remarkable performance across vision-language tasks. Recent advancements allow these models to process multiple images as inputs. However, the vulnerabilities of multi-image MLLMs remain unexplored. Existing adversarial attacks focus on single-image settings and often assume a white-box threat model, which is impractical in many real-world scenarios. This paper introduces LAMP, a black-box method for learning Universal Adversarial Perturbations (UAPs) targeting multi-image MLLMs. LAMP applies an attention-based constraint that prevents the model from effectively aggregating information across images. LAMP also introduces a novel cross-image contagious constraint that forces perturbed tokens to influence clean tokens, spreading adversarial effects without requiring all inputs to be modified. Additionally, an index-attention suppression loss enables a robust position-invariant attack. Experimental results show that LAMP outperforms SOTA baselines and achieves the highest attack success rates across multiple vision-language tasks and models.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 32 of its 45 references that are in the map. First posted on arXiv 29 Jan 2026.

Builds on

32 of 45 references.

Closest ancestors · 4

Same direction, newest first

  1. 2024Exploiting Vision Encoder Vulnerabilities for Universal Adversarial Perturbations on Large Vision-Language Models11 Dec
  2. 2024Efficient and Effective Universal Adversarial Attack against Vision-Language Pre-training Models15 Oct
  3. 2023Set-level Guidance Attack: Boosting Adversarial Transferability of Vision-Language Pre-training Models26 Jul
  4. 2022Towards Adversarial Attack on Vision-Language Pre-training Models19 Jun

Other references · 28

  1. 2021Learning Transferable Visual Models From Natural Language Supervisioncited by 33,456
  2. 2014Microsoft COCO: Common Objects in Contextcited by 17,498
  3. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  4. 2013Intriguing properties of neural networkscited by 5,114
  5. 2023MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilitiescited by 1,221
  6. 2019OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledgecited by 1,114
  7. 2016Universal adversarial perturbationscited by 774
  8. 2021NExT-QA:Next Phase of Question-Answering to Explaining Temporal Actionscited by 735
  9. 2019Universal Adversarial Triggers for Attacking and Analyzing NLPcited by 709
  10. 2018Improving Transferability of Adversarial Examples with Input Diversitycited by 591
  11. 2024BLINK: Multimodal Large Language Models Can See but Not Perceivecited by 529
  12. 2018A Corpus for Reasoning About Natural Language Grounded in Photographscited by 457
16 more
  1. 2023Q-Bench: A Benchmark for General-Purpose Foundation Models on Low-level Visioncited by 247
  2. 2024MANTIS: Interleaved Multi-Image Instruction Tuningcited by 212
  3. 2020Skip Connections Matter: On the Transferability of Adversarial Examples Generated with ResNetscited by 165
  4. 2019Enhancing Adversarial Example Transferability with an Intermediate Level Attackcited by 139
  5. 2019Cycle-Consistency for Robust Visual Question Answeringcited by 105
  6. 2023AdvCLIP: Downstream-agnostic Adversarial Examples in Multimodal Contrastive Learningcited by 101
  7. 2021Towards Transferable Adversarial Attacks on Vision Transformerscited by 80
  8. 2021PatchCleanser: Certifiably Robust Defense against Adversarial Patches for Any Image Classifiercited by 63
  9. 2020Perturbing Across the Feature Hierarchy to Improve Standard and Strict Blackbox Attack Transferabilitycited by 57
  10. 2020Transferable Perturbations of Deep Feature Distributionscited by 51
  11. 2017Art of singular vectors and universal adversarial perturbationscited by 49
  12. 2021Random Noise Defense Against Query-Based Black-Box Attackscited by 46
  13. 2020Blacklight: Scalable Defense for Neural Networks against Query-Based Black-Box Attackscited by 34
  14. 2022Boosting the Transferability of Adversarial Attacks with Global Momentum Initializationcited by 29
  15. 2020Adversarial Robustness through Bias Variance Decomposition: A New Perspective for Federated Learningcited by 13
  16. 2019On Single Source Robustness in Deep Fusion Modelscited by 12

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026Universal Adversarial Attacks against Closed-Source MLLMs via Target-View Routed Meta Optimization
  2. 2026Breaking the weakest link to evade vision language models
  3. 2025Transferable Adversarial Attacks on Black-Box Vision-Language Models
  4. 2024Universal Adversarial Perturbations for Vision-Language Pre-trained Models
  5. 2024Multimodal Adversarial Defense for Vision-Language Models by Leveraging One-To-Many Relationships
  6. 2026Unveiling the Fragility of Vision-Language Models: Multi-Modal Adversarial Synergy via Texture-Constrained Perturbations and Cross-Modal Optimization
  7. 2024One Perturbation is Enough: On Generating Universal Adversarial Perturbations against Vision-Language Pre-training Models
  8. 2026Crafting Adversarial Inputs for Large Vision-Language Models Using Black-Box Optimization