ActiveVLA: Injecting Active Perception into Vision-Language-Action Models for Precise 3D Robotic Manipulation Paper • 2601.08325 • Published Jan 13 • 1