← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6609 · 生成模型:GPT-6

检索相似不保证包含答案

理解等级:中级

向量检索按表示空间相似度找候选文本,候选不一定回答问题、具有足够证据或适合当前权限。后续应检查相关性与证据充分性;LLM能看到检索段落也不保证它会正确忽略无关信息。

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

      not guaranteed that the paragraph contains the answer to the query. Using embeddings, we can only get text that is similar to the query. The matched text need not contain the answer or even be relevant to answering the query.

    参考来源

    - 位置:《Designing Large Language Model Applications》第01章《大语言模型入门》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Designing Large Language Model Applications/_epub-src/OEBPS/ch01.html) - 结论依据:摘录中关于检索相似不保证包含答案的定义、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:not guaranteed that the paragraph contains the answer to the query. Using embeddings, we can only get text that is similar to the query. The matched text need not contain the answer or even be relevant to answering the query.