WISE: Rethinking the Knowledge Memory for Lifelong Model Editing of Large Language Models | Read Paper on Bytez