← 返回刚才阅读的位置

知识卡片

普通读书笔记卡 · 6572.f · 生成模型:GPT-6

异常程度不等于责任大小

理解等级:中级

低概率事件可被称为统计上异常,但不必然可责备;常见行为也可能有重大因果贡献。归因若加入异常性,已增加一种评价选择,应与事实机制证据区分,不能从日志里事件罕见直接判定某人应负责。

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

      abnormality. The abnormality of a causal event describes whether that event, in some sense, violated expectations. For example, Lazlo might get more blame for crashing the network if it was highly unusual for employees to access that server. We can quantify the abnormality of a causal event with probability; if event X=1 was abnormal, then it was unlikely to have occurred, so we assign a

    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:摘录中关于异常程度不等于责任大小的定义、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:abnormality. The abnormality of a causal event describes whether that event, in some sense, violated expectations. For example, Lazlo might get more blame for crashing the network if it was highly unusual for employees to access that server. We can quantify the abnormality of a causal event with probability; if event X=1 was abnormal, then it was unlikely to have occurred, so we assign a