mlsanityData to 7 Oct 2026

Hijacking Context in Large Multi-modal Models

arXiv 2312.07553 ↗First posted 7 Dec 2023cs.AI cs.CLcited by 6

On the map: Multimodal LLMs / vision-language models → MLLM training and visual instruction tuning

Abstract (arXiv)

Recently, Large Multi-modal Models (LMMs) have demonstrated their ability to understand the visual contents of images given the instructions regarding the images. Built upon the Large Language Models (LLMs), LMMs also inherit their abilities and characteristics such as in-context learning where a coherent sequence of images and texts are given as the input prompt. However, we identify a new limitation of off-the-shelf LMMs where a small fraction of incoherent images or text descriptions mislead LMMs to only generate biased output about the hijacked context, not the originally intended context. To address this, we propose a pre-filtering method that removes irrelevant contexts via GPT-4V, based on its robustness towards distribution shift within the contexts. We further investigate whether replacing the hijacked visual and textual contexts with the correlated ones via GPT-4V and text-to-image models can help yield coherent responses.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Placed in MLLM training and visual instruction tuning, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 20 of its 27 references that are in the map. First posted on arXiv 7 Dec 2023.

Builds on

20 of 27 references.

References · 20

  1. 2020Language Models are Few-Shot Learnerscited by 32,199
  2. 2023GPT-4 Technical Reportcited by 18,132
  3. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
  4. 2023Visual Instruction Tuningcited by 7,792
  5. 2017Deep reinforcement learning from human preferencescited by 3,846
  6. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  7. 2023Lost in the Middle: How Language Models Use Long Contextscited by 3,383
  8. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,898
  9. 2023MiniGPT-4: Enhancing Vision-Language Understanding with Advanced Large Language Modelscited by 2,689
  10. 2019Transformer-XL: Attentive Language Models Beyond a Fixed-Length Contextcited by 1,830
  11. 2022Rethinking the Role of Demonstrations: What Makes In-Context Learning Work?cited by 1,413
  12. 2022Solving Quantitative Reasoning Problems with Language Modelscited by 1,359
8 more
  1. 2023Large Language Models Can Be Easily Distracted by Irrelevant Contextcited by 820
  2. 2021MetaICL: Learning to Learn In Contextcited by 450
  3. 2023Visual Adversarial Examples Jailbreak Aligned Large Language Modelscited by 307
  4. 2016Visual Storytellingcited by 176
  5. 2016Exploring the Limits of Language Modelingcited by 172
  6. 2023On the Adversarial Robustness of Multi-Modal Foundation Modelscited by 142
  7. 2023Adversarial Demonstration Attacks on Large Language Modelscited by 58
  8. 2023Hijacking Large Language Models via Adversarial In-Context Learningcited by 41

Built on by

Cited by year

Citing papers per year: 2024: 2; 2025: 4.

6 papers.

  1. 2025Why Language Models Hallucinatecited by 205
  2. 2025Why and How LLMs Hallucinate: Connecting the Dots with Subsequence Associationscited by 14
  3. 2024GenTel-Safe: A Unified Benchmark and Shielding Framework for Defending Against Prompt Injection Attackscited by 7
  4. 2024Towards Action Hijacking of Large Language Model-based Agentcited by 4
  5. 2025On the Robustness of Transformers against Context Hijacking for Linear Classificationcited by 4
  6. 2025Towards LLM-Centric Multimodal Fusion: A Survey on Integration Strategies and Techniquescited by 2

Similar, not cited

Close in wording, no citation either way

  1. 2024AIM: Let Any Multi-modal Large Language Models Embrace Efficient In-Context Learning
  2. 2024Browse and Concentrate: Comprehending Multimodal Content via prior-LLM Context Fusion
  3. 2025Long-VITA: Scaling Large Multi-modal Models to 1 Million Tokens with Leading Short-Context Accuracy
  4. 2024Efficient Large Multi-modal Models via Visual Context Compression
  5. 2025HaploVL: A Single-Transformer Baseline for Multi-Modal Understanding
  6. 2023Eliciting In-Context Learning in Vision-Language Models for Videos Through Curated Data Distributional Properties
  7. 2026Visual-Noise Guided In-Context Distillation for Multimodal Large Language Model Unlearning
  8. 2024CoF: Coarse to Fine-Grained Image Understanding for Multi-modal Large Language Models