Analyzing Position Bias in Zero-Shot Abstractive Summarization with Large Language Models

Position bias in large language models affects the quality of zero-shot abstractive summarization.