mlsanityData to 14 Oct 2026

Self-adaptive Dataset Construction for Real-World Multimodal Safety Scenarios

arXiv 2509.04403 ↗First posted 4 Sep 2025cs.CV cs.CL cs.CRcited by 2

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

Abstract (arXiv)

Multimodal large language models (MLLMs) are rapidly evolving, presenting increasingly complex safety challenges. However, current dataset construction methods, which are risk-oriented, fail to cover the growing complexity of real-world multimodal safety scenarios (RMS). And due to the lack of a unified evaluation metric, their overall effectiveness remains unproven. This paper introduces a novel image-oriented self-adaptive dataset construction method for RMS, which starts with images and end constructing paired text and guidance responses. Using the image-oriented method, we automatically generate an RMS dataset comprising 35k image-text pairs with guidance responses. Additionally, we introduce a standardized safety dataset evaluation metric: fine-tuning a safety judge model and evaluating its capabilities on other safety datasets.Extensive experiments on various tasks demonstrate the effectiveness of the proposed image-oriented pipeline. The results confirm the scalability and effectiveness of the image-oriented approach, offering a new perspective for the construction of real-world multimodal safety datasets. The dataset is presented at https://huggingface.co/datasets/NewCityLetter/RMS2/tree/main.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in MLLM jailbreaks and adversarial robustness, one of 16 directions in Multimodal LLMs / vision-language models. Builds on 19 of its 32 references that are in the map. First posted on arXiv 4 Sep 2025.

Builds on

19 of 32 references.

Closest ancestors · 6

Same direction, newest first

  1. 2025Visual Contextual Attack: Jailbreaking MLLMs with Image-Driven Context Injection3 Jul
  2. 2024VLSBench: Unveiling Visual Leakage in Multimodal Safety29 Nov
  3. 2024Llama Guard 3 Vision: Safeguarding Human-AI Image Understanding Conversations15 Nov
  4. 2024Learning To See But Forgetting To Follow: Visual Instruction Tuning Makes LLMs More Prone To Jailbreak Attacks7 May
  5. 2023FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual Prompts9 Nov
  6. 2023Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models26 Jul

Other references · 13

  1. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  2. 2023Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyondcited by 2,224
  3. 2023Jailbroken: How Does LLM Safety Training Fail?cited by 1,651
  4. 2024Expanding Performance Boundaries of Open-Source Multimodal Models with Model, Data, and Test-Time Scalingcited by 1,599
  5. 2022The Carbon Footprint of Machine Learning Training Will Plateau, Then Shrinkcited by 184
  6. 2023How Many Unicorns Are in This Image? A Safety Evaluation Benchmark for Vision LLMscited by 108
  7. 2023DRESS: Instructing Large Vision-Language Models to Align and Interact with Humans via Natural Language Feedbackcited by 103
  8. 2024InferAligner: Inference-Time Alignment for Harmlessness through Cross-Model Guidancecited by 83
  9. 2024GOAT-Bench: Safety Insights to Large Multimodal Models through Meme-Based Social Abusecited by 52
  10. 2025T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generationcited by 48
  11. 2024Assessment of Multimodal Large Language Models in Alignment with Human Valuescited by 36
  12. 2024Red Teaming GPT-4V: Are GPT-4V Safe Against Uni/Multi-Modal Jailbreak Attacks?cited by 34
1 more
  1. 2023Large Language Models as Automated Aligners for benchmarking Vision-Language Modelscited by 10

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 1.

2 papers.

  1. 2025Contextual Image Attack: How Visual Context Exposes Multimodal Safety Vulnerabilitiescited by 3
  2. 2026MCBench: A Multicontext Safety Assessment Benchmark for Omni Large Language Modelscited by 1

Similar, not cited

Close in wording, no citation either way

  1. 2025SDEval: Safety Dynamic Evaluation for Multimodal Large Language Models
  2. 2026When Helpers Become Hazards: A Benchmark for Analyzing Multimodal LLM-Powered Safety in Daily Life
  3. 2025Can't See the Forest for the Trees: Benchmarking Multimodal Safety Awareness for Multimodal LLMs
  4. 2024Multimodal Situational Safety
  5. 2024MLLMGuard: A Multi-dimensional Safety Evaluation Suite for Multimodal Large Language Models
  6. 2026KSAFE-MM: A Multimodal Safety Benchmark via Localized Contextualization for Korean Cultural Risks
  7. 2024Safety of Multimodal Large Language Models on Images and Texts
  8. 2025Multimodal Large Language Models for Enhanced Traffic Safety: A Comprehensive Review and Future Trends