Detect and Suppress: A Mechanistic Defense against Adversarial Patches in VLA Models

arXiv:2610.03498 · cs.RO, cs.AI · Submitted 2026-10-02 · Read on arXiv

cs.RO, cs.AI

Submitted: 2026-10-02

Updated: 2026-10-02

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Terminology

• VLA Models — 1× in this paper · explained in 3 episodes

Sources

• Model-agnostic Adversarial Attack and Defense for Vision-Language-Action Models
• Event-Grounded Sparse Autoencoders for Vision-Language-Action Policies
• SmolVLA: A Vision-Language-Action Model for Affordable and Efficient Robotics
• Adversarial Patch
• DPatch: An Adversarial Patch Attack on Object Detectors
• When Alignment Fails: Multimodal Adversarial Attacks on Vision-Language-Action Models
• Steering CLIP's vision transformer with sparse autoencoders

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• ARC: A Reasoning Recipe for Robot Foundation Models
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies