SafeAtlas-VL introduces a large multimodal safety dataset with 1.5 million instances, rating image, request, and response risks on a five‑level ordinal scale across 15 harm categories and 55 subcategories. The accompanying SafeAtlas‑Bench provides 5,000 held‑out cases for evaluating ordinal predictions and continuous risk scores. Models trained on this data, including an 8B Guard model, achieve state‑of‑the‑art performance, outperforming prior benchmarks by about 4% in F1 score.
By Zongrui Wang, Xiangyang Zhu, Sicheng Wang, Han Wang, Dingyi Rong, Zeyu Zhang, Chunyi Li, Yue Shi, Kaiwei Zhang, Zicheng Zhang, Yuan Tian, Qi Jia, Yan Teng, Wei Sun, Ning Liu, Guangtao Zhai
Multi-modal Large Language Models (MLLMs) achieve strong performance on vision-language tasks, but incorporating visual inputs through a vision encoder (e. g.
While Large Vision-Language Models (LVLMs), represented by LLaVA and GPT-4V, have demonstrated remarkable capabilities, their visual inputs remain vulnerable to adversarial attacks, posing significant security risks. Existing defense methods predominantly target single-task scenarios (e.
arXiv:2607. 18958v1 Announce Type: cross Abstract: While Large Vision-Language Models (LVLMs), represented by LLaVA and GPT-4V, have demonstrated remarkable capabilities, their visual inputs remain vulnerable to adversarial attacks, posing significant security risks.
By Sibo Wang, Jie Zhang, Shiguang Shan, Xilin Chen, Wen Gao
The paper investigates how Vision Language Models (VLMs) can be fooled by tiny, human‑imperceptible changes to images. It introduces a gradient‑based attack that targets only the vision encoder, reducing computational cost while still effectively disrupting both untargeted and targeted multimodal interpretations. Experiments on open‑source VLMs such as Qwen2.5‑VL, Granite‑Vision, FastVLM, and Phi‑3.5‑Vision demonstrate that these small perturbations can dramatically alter the models’ textual outputs.
The paper investigates how Vision Language Models (VLMs) can be fooled by small, human‑imperceptible changes to images. It introduces a gradient‑based attack that targets only the vision encoder, reducing computational cost while still effectively disrupting both untargeted and targeted multimodal alignment. Experiments on open‑source VLMs such as Qwen2.5‑VL, Granite‑Vision, FastVLM, and Phi‑3.5‑Vision demonstrate that these perturbations can significantly alter the models’ textual outputs.
By Ilan Zini, Boussad Addad, Katarzyna Kapusta
Vision-language models (VLMs) are increasingly deployed in consumer, medical, financial, and enterprise applications. This broad deployment expands the safety surface: risks can arise from multimodal question answering, assistant responses, and cross-modal composition, while moderation policies may vary across products, regions, and deployment stages.
arXiv:2606. 24388v1 Announce Type: new Abstract: We introduce a large-scale, open-source dataset of pre-generated adversarial attacks for vision-language models (VLMs).
By Simone Gallivanone, Hossein Khodadadi, Mauro Dore, Mauro Medda, Nicola Franco
MM-IFEval-Pro is a new multilingual benchmark for evaluating instruction-following in vision-language models, covering both Chinese and English tasks. It includes 4 major task categories, 24 subcategories, and 8 instruction categories with 52 subcategories, each sample featuring an average of 3.0 constraints to mimic complex instruction scenarios. A reinforcement-learning training set with Chinese and adversarial instructions improves model performance on MM-IFEval-Pro and transfers well to other multimodal benchmarks, showing strong cross-task and cross-language generalization.
By Changming Xiao, Zhenliang Ni, Jinhui He, Han Shu, Jie Hu
arXiv:2609.39688v1 Announce Type: cross
Abstract: Multimodal encoders such as CLIP underlie many downstream systems, but their web-scale training data embed harmful associations that safety alignment...
By Tobia Poppi, Silvia Cappelletti, Samuele Poppi, Marcella Cornia, Lorenzo Baraldi, Diego Garcia-Olano, Rita Cucchiara
arXiv:2609.22234v1 Announce Type: cross
Abstract: Instruction hierarchy (IH) alignment teaches language models to prioritize higher-level instructions when inputs conflict. While studied primarily in...
By Nicholas Sansoterra, Zishuo Zheng, Sachin Kumar
arXiv:2602. 03300v2 Announce Type: replace-cross Abstract: In this work, we aim to develop effective data synthesis techniques that autonomously synthesize multimodal training data for enhancing MLLMs in solving complex real-world tasks.
By Jingyi Zhang, Tianyi Lin, Huanjin Yao, Xiang Lan, Shunyu Liu, Jiaxing Huang