Sequential statistical inference for Large Language Models: Representation, validity, and monitoring
大语言模型的序贯统计推断:表示、有效性与监控
机构 * H. Milton Stewart School of Industrial and Systems Engineering, Georgia Institute of Technology(佐治亚理工学院工业与系统工程系)
专题命中 幻觉与事实性 :trustworthy(abstract);分类 cs.LG
AI总结 本文提出将序贯统计推断应用于大语言模型可信赖性,围绕表示、有效性和监控三个任务展开,将LLM交互视为依赖随机过程,提供不确定性保证并检测行为变化。
Comments This article was prepared for a invited discussion in The American Statistician