The paper is accepted at IJCAI 2024 ð [ Memorizing Documents with Guidance in Large Language Models ]