Research has shown that large language models (LLMs) tend to overemphasize information at the beginning and end of a document ...