Object-Centric Prompt-Driven Vision-Language-Action Model for Robotic Manipulation | Read Paper on Bytez