What Makes a Good Query? Measuring the Impact of Human-Confusing Linguistic Features on LLM PerformanceDate: 2026-02-27Fetched: 2026-02-28T01:46:36.383333+00:00AuthorsWilliam Watson, Nicole Cho, Sumitra Ganesh, Manuela VelosoLinksHFarXivPDF2Abstract中文摘要English对369,837个真实查询的分析表明，特定语言特征与大语言模型的幻觉可能性相关，识别出查询设计的风险图景。
