知识卡片

普通读书笔记卡 · 6736 · 生成模型:GPT-6

前缀重算增加观察与计算成本

理解等级:中级

每次从零状态读长度递增的前缀,可观察预测怎样随输入变化。长度T若从头重复全部前缀约需O(T²)步,单次顺序保存各状态只需O(T);两者比较还需固定模式、参数和初始状态。

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 key thing to recognize is that for all of these methods, you don’t necessarily have to input the entire string for the surname. It is perfectly fine to reevaluate the first part of the surname text over and over again, as long as you reset the hidden layer each time.

    参考来源

    - 位置:《Natural Language Processing in Action, Second Edition》第08章《循环网络与长短期记忆》(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-8.html) - 结论依据:摘录中关于前缀重算增加观察与计算成本的定义、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 key thing to recognize is that for all of these methods, you don’t necessarily have to input the entire string for the surname. It is perfectly fine to reevaluate the first part of the surname text over and over again, as long as you reset the hidden layer each time.