AI Paper Atlas

Research map / Vision-language and multimodal / Human and embodied modeling

Attacking Gray-Box Large Vision-Language Models with Adaptive SVD-Structured Adversarial Alignment

Daizong Liu, Xiaowen Cai, Junhao Dong, Zhongliang Guo, Xiaoye Qu, Runwei Guan, Xiang Fang, Dengpan Ye

ICML 2026 · Vision-language and multimodal › Human and embodied modeling · approach: Scene graph generation

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add coarse alignment to the no-alignment baseline (ablation)Qwen2.5-VL-3B attack · COCOASR: 0.0 -> 69.8Improved
add fine-grained alignment to the no-alignment baseline (ablation)Qwen2.5-VL-3B attack · COCOASR: 0.0 -> 78.5Improved
add fine-grained alignment on top of coarse alignment (ablation)Qwen2.5-VL-3B attack · COCOASR: 69.8 -> 86.5Improved
add coarse alignment on top of fine-grained alignment (ablation)Qwen2.5-VL-3B attack · COCOASR: 78.5 -> 86.5Improved
add coarse alignment on top of fine-grained alignment (ablation)LLaVa-1.5-7B attack · COCOASR: 75.2 -> 84.7Improved
add fine-grained alignment on top of coarse alignment (ablation)LLaVa-1.5-7B attack · COCOASR: 63.7 -> 84.7Improved
use projection-based cost C_Pi in OT (ablation)Ours · COCO, Qwen2.5-VL-3BASR: 82.4 -> 86.5Improved
use projection-based cost C_Pi in OT (ablation)Ours · COCO, Gemma-3-12BASR: 71.6 -> 74.2Improved
use projection-based cost C_Pi in OT (ablation)Ours · COCO, LLaVa-1.6-7BASR: 75.1 -> 78.5Improved
replace CLIP-ViT text encoder with Sentence-Bert (ablation)Ours · COCO, Qwen2.5-VL-3BASR: 86.5 -> 87.3No clear effect
replace CLIP-ViT text encoder with Sentence-Bert (ablation)Ours · COCO, LLaVa-1.5-7BASR: 84.7 -> 85.1No clear effect
increase number K of principal components in projection (ablation)Ours · COCO, Qwen2.5-VL-3B and LLaVa-1.5-7BsimilarityImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy