3D Question Answering via only 2D Vision-Language Models | Read Paper on Bytez