arXiv AI September 10, 2026 By Dongsu Song, DaeYun GO, Jay Hoon Jung

Discovering Natural Transformation Vulnerabilities in Black-Box Vision Models

The paper introduces Adversarial Scenario Attack (ASA), a query‑based black‑box method that discovers natural transformation vulnerabilities in vision models by exploring background, weather, and material/color edits via a multimodal language model and a text‑guided generative editor. ASA outperforms previous query‑based generative attacks on ImageNet classifiers, achieving higher success rates with fewer queries while maintaining perceptual quality. The approach also shows image‑level and prompt‑level transferability, indicating reusable vulnerabilities across models and images.

Machine-generated by The Flow from the publisher's headline and feed description — not written or checked by a human. The full article lives at arXiv AI.

arXiv Computer Vision

1d ago

BackdoorVLM: A Benchmark for Backdoor Attacks and Defenses on Vision-Language Models

arXiv:2511.18921v2 Announce Type: replace Abstract: Backdoor attacks undermine the reliability and trustworthiness of machine learning systems by injecting hidden behaviors that can be maliciously ac...

By Juncheng Li, Yige Li, Hanxun Huang, Yunhao Chen, Xin Wang, Yixu Wang, Xingjun Ma, Yu-Gang Jiang

llms nlp multimodal benchmarks safety

arXiv Machine Learning

Jun 3

BYORn: Bootstrap Your Own Responses to Defend Large Vision-Language Models Against Backdoor Attacks

arXiv:2606. 02947v1 Announce Type: new Abstract: Supervised fine-tuning is the predominant approach for adapting autoregressive vision-language models to downstream tasks.

By Ivan Saboli\'c, Marin Or\v{s}i\'c, Josip \v{S}ari\'c, Sven Lon\v{c}ari\'c

llms fine-tuning multimodal

arXiv AI

Aug 19

DiSCO: Defending text-to-image generation through distribution-guided contrastive prompt optimization

DiSCO is a zero‑shot, black‑box defense for text‑to‑image models that operates solely at the prompt level. It expands prompts with a distribution‑guided suffix using beam search and contrastive scoring against safe and unsafe image pools generated by the target model, iteratively refining until safe content is produced. The method improves safety on the I2P benchmark under various red‑teaming attacks, reducing attack success rates by 37.7% and 25.13% while preserving semantic fidelity and image coherence.

By Tong Zhang, Motasem Alfarra, Carlos Hinojosa, Christos Louizos, Bernard Ghanem

llms diffusion fine-tuning multimodal benchmarks safety

arXiv AI

Jun 29

When the Prompt Becomes Visual: Vision-Centric Jailbreak Attacks for Large Image Editing Models

arXiv:2602. 10179v2 Announce Type: replace-cross Abstract: Recent advances in large image editing models have shifted the paradigm from text-driven instructions to vision-prompt editing, where user intent is inferred directly from visual inputs such as marks, arrows, and visual-text prompts.

By Jiacheng Hou, Yining Sun, Ruochong Jin, Haochen Han, Fangming Liu, Wai Kin Victor Chan, Alex Jinpeng Wang

llms multimodal benchmarks safety

arXiv Machine Learning

Aug 14

Adjustable Text-Guided Backdoor Attacks with Natural-Word Triggers on Multimodal Pretrained Models

arXiv:2604. 05809v2 Announce Type: replace-cross Abstract: This paper presents Text-Guided Backdoor (TGB), an adjustable backdoor attack against multimodal pretrained models that uses natural-word triggers, namely words that can naturally occur in ordinary textual inputs.

By Yiyang Zhang, Chaojian Yu, Ziming Hong, Yuanjie Shao, Qinmu Peng, Tongliang Liu, Xinge You

nlp multimodal safety

Hugging Face Trending Papers

Aug 19

Breaking the weakest link to evade vision language models

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.

llms robotics multimodal safety