Research map / Adversarial robustness and security / Black-box and transfer attacks
Dual Feature-Relational Alignment for Transferable Targeted Attacks on MLLMs
Dongshen Han, Xiaorong Tian, Sheng Zheng, Zishan Huang, Boyu Deng, Wei Dong, Zhijie Li, Yang Yang, Chaoning Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes a dual feature-relational alignment attack to craft transferable targeted adversarial examples against black-box MLLMs from surrogate models.
Problem
Targeted transferable adversarial attacks on multimodal large language models are hard because perturbations must inject target semantics and generalize from surrogate models to unseen black-box MLLMs; existing global alignment or local matching overfits to surrogate specifics.
Approach
Aligns feature relations (rather than only coarse global features or unstable local features) between adversarial and target inputs to encourage transferable target semantics; details beyond the abstract are not available.
Closest papers by meaning
- Adversarial Attacks against Closed-Source MLLMs via Feature Optimal Alignment (NeurIPS 2025 · similarity 0.89)
- Semantic Representation Attack against Aligned Large Language Models (NeurIPS 2025 · similarity 0.88)
- Stage-wise Attention-Guided Region Sequencing for Adversarial Attacks on Large Vision-Language Models (NeurIPS 2026 · similarity 0.87)
- A Frustratingly Simple Yet Highly Effective Attack Baseline: Over 90% Success Rate Against the Strong Black-box Models of GPT-4.5/4o/o1 (NeurIPS 2025 · similarity 0.87)
- Chain of Attack: On the Robustness of Vision-Language Models Against Transfer-Based Adversarial Attacks (CVPR 2025 · similarity 0.86)
- Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models (ICLR 2024 · similarity 0.86)
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.
