← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 9114.f · 生成模型:GPT-6

向量维数与张量阶数应区分

理解等级:初级

“128维特征”指向量有128个坐标,而“四维张量”通常指数组有四个轴,各轴还各有长度。两者使用同一词但含义不同。描述模型输入、批次和特征时,给出具体shape及轴名能避免把特征数量误当数组阶数。

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

      The word “dimension” has two usages in the computational sciences. The first usage is as a coordinate in a multivariate data structure, for example, “the i-th dimension of a vector” or “a 128-dimensional feature space.” The second usage is as the shape of a multidimensional array, as in “a 4D tensor.” We will use both these meanings in this book and we hope the usage

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》符号约定专题《符号与矩阵微积分约定》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c000b.xhtml) - 结论依据:摘录中关于向量维数与张量阶数应区分的定义、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:The word “dimension” has two usages in the computational sciences. The first usage is as a coordinate in a multivariate data structure, for example, “the i-th dimension of a vector” or “a 128-dimensional feature space.” The second usage is as the shape of a multidimensional array, as in “a 4D tensor.” We will use both these meanings in this book and we hope the usage