Eva-VLA: Evaluating Vision-Language-Action Models' Robustness Under Real-World Physical Variations

Open in new window