知识卡片

普通读书笔记卡 · 8985.b · 生成模型:GPT-6

链式分解不要求条件项独立

理解等级:中级

任意联合分布可按固定顺序写为p(x1)∏i>1p(xi|x)。从每个归一化条件分布依次抽样即可产生对应联合分布样本,原因是概率链式法则,不是这些条件分布或像素相互独立。若仅用局部前缀,则额外施加了上下文限制。

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

      So we can model the probability of each subsequent pixel given the preceding pixels. To generate an image we can sample a value for the first pixel, then sample the second given the first, then the third given the first and second, and so forth. But is this a valid model of p(X) = p(x0, …, xn), the probability distribution of the full set of

    参考来源

    - 位置:《Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)》第32章《生成模型》(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/c032.xhtml) - 结论依据:摘录中关于链式分解不要求条件项独立的定义、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:So we can model the probability of each subsequent pixel given the preceding pixels. To generate an image we can sample a value for the first pixel, then sample the second given the first, then the third given the first and second, and so forth. But is this a valid model of p(X) = p(x0, …, xn), the probability distribution of the full set of