← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 8696 · 生成模型:GPT-6

摘要质量需拆成独立标准

理解等级:中级

摘要至少要检查忠实性、具体性、相关性、完整性、重复、连贯性及格式结构。词面相似度只能覆盖部分维度;总分可能掩盖关键数值错误或遗漏,应按业务需求分别说明必要标准。

这张卡片看不懂?

已有反馈 · 0 条

    还没有人提交反馈。

    在整书笔记中的位置

    原书出处 · 1 处

    • · 查看原书依据

      Factuality The summary is factually correct and does not make incorrect assumptions or conclusions from the source text. Specificity The summary doesn’t oversummarize; it avoids being generic and provides specific details, whether numbers or named entities. Relevance Also called precision, this is calculated as the percentage of sentences in the summary that are deemed relevant and thus merit inclusion in the summary. Completeness Also called

    参考来源

    - 位置:《Designing Large Language Model Applications》第10章《外部工具与智能体》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch10.html) - 结论依据:摘录中关于摘要质量需拆成独立标准的定义、比较或条件支撑卡片;原书表述不严谨的范围已在覆盖记录逐项限定,不把案例结果外推为普遍结论。 - 原始内容:Factuality The summary is factually correct and does not make incorrect assumptions or conclusions from the source text. Specificity The summary doesn’t oversummarize; it avoids being generic and provides specific details, whether numbers or named entities. Relevance Also called precision, this is calculated as the percentage of sentences in the summary that are deemed relevant and thus merit inclusion in the summary. Completeness Also called