📄 中文摘要
摘要是自然语言处理（NLP）中的核心任务。随着大型语言模型（LLMs）的进步以及大上下文窗口的引入，使得在单个提示中处理整本书成为可能。同时，对于知名书籍，LLMs可以仅基于训练期间获得的内部知识生成摘要。这引发了一些重要问题：基于内部记忆生成的摘要与基于完整文本的摘要相比如何？即使模型提供了书籍作为输入，先前的知识是否会影响摘要？该研究通过实验评估了使用最先进的LLMs进行书籍摘要的效果，比较了生成的知名书籍摘要。
大型语言模型与书籍摘要：阅读与记忆，哪种更优？
📄 English Summary
Large Language Models and Book Summarization: Reading or Remembering, Which Is Better?
Summarization is a fundamental task in Natural Language Processing (NLP). Recent advancements in Large Language Models (LLMs) and the introduction of large context windows allow for the processing of entire books in a single prompt. Additionally, for well-known books, LLMs can generate summaries based solely on internal knowledge acquired during training. This raises critical questions: How do summaries generated from internal memory compare to those derived from the full text? Does prior knowledge influence summaries even when the model is given the book as input? This study conducts an experimental evaluation of book summarization using state-of-the-art LLMs, comparing the summaries produced for well-known books.