知识卡片

普通读书笔记卡 · 6697 · 生成模型:GPT-6

向量组合产生候选而非新事实

理解等级:中级

将若干词向量相加,可形成混合方向并检索附近词项;相减可尝试削弱某种方向。结果依赖范数、归一化及词表,不自动创造新词或证明属性组合真实;加上Chief与Engineer的近邻也不是职业晋升路线。

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

      Another nice thing about word embeddings is that you can combine them any way you like to create new words! You can add the meanings of the words together to try to find a single word that captures the meaning of the two words you added together:

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第06章《用词嵌入推理》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Natural Language Processing in Action, Second Edition/_epub-src/OEBPS/Text/chapter-6.html) - 结论依据:摘录中关于向量组合产生候选而非新事实的定义、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:Another nice thing about word embeddings is that you can combine them any way you like to create new words! You can add the meanings of the words together to try to find a single word that captures the meaning of the two words you added together: