Approximate Domain Unlearning for Vision-Language Models | Read Paper on Bytez