Attention-Guided Patch-Wise Sparse Adversarial Attacks on Vision-Language-Action Models

作者：ADVLA: ATTENTION-GUIDED, Wei Tao, Xi Xiao, Qianpu Sun, Yuxin Zheng, Wentao Mo, Peiqiang Wang, Nan Zhang · 单位：Shenzhen International Graduate School, Tsinghua University, Shenzhen, China, Huazhong University of Science and Technology, Wuhan, China, Ping An Technology, Shenzhen, China · 会议/期刊：arXiv · 日期：2025-11-26 · 来源：Low-Level Learning-Based Action Modelling / Input Modelling / 2D Vision Language Action Models with Robustness / Attack