A Closer Look at the Few-Shot Adaptation of Large Vision-Language Models | Read Paper on Bytez