← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6574.e · 生成模型:GPT-6

世界简单性只是特定操作后果

理解等级:中级

对同一图执行理想干预只删入边,因此相应干预图的边是原图子集。不同世界可有不同目标,互相未必嵌套;状态政策还可能引入新依赖。不能把“假想世界更简单”当作通用可识别或数值稳定保证。

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

      8.4.5 Rule of thumb: Hypothetical worlds should be simpler Consider again the endogenous nodes in our online gaming example in figure 8.16. Notice that, in this example, the two worlds have the same sets of endogenous nodes, and the edges in the hypothetical world are a subset of the edges of those in the actual world. In other words, the possible world where we do

    参考来源

    - 位置:《Causal AI》第08章《反事实与平行世界》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-8.xhtml) - 结论依据:摘录中关于世界简单性只是特定操作后果的定义、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:8.4.5 Rule of thumb: Hypothetical worlds should be simpler Consider again the endogenous nodes in our online gaming example in figure 8.16. Notice that, in this example, the two worlds have the same sets of endogenous nodes, and the edges in the hypothetical world are a subset of the edges of those in the actual world. In other words, the possible world where we do