知识卡片
潜在编码只是未建模原因替身
VAE引入低维Z承载未被数字和书写标签表达的图像变化。可在假设模型中赋予它潜在原因替身的解释,但重建训练并不证明其各坐标对应真实可独立干预的因素。表示有用与因果因素已恢复需要不同证据。
这张卡片看不懂?
这张卡片已通过人工审查,不再接受失败反馈。
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 variations in the handwritten digits due to different writers and motor movements as they wrote. With that in mind, we can view P(X| digit, is-handwritten, Z) as the causal Markov kernel of X. That said, it is important to remember that the representation we learn for Z is a stand-in for latent causes and is not the same as learning the actual latent causes.
参考来源
- 位置:《Causal AI》第05章《连接因果与深度学习》(来源项目:a19c7933-4eda-4176-aebd-30fe7e1ad292;源文件:extracted/Causal AI/_epub-src/OEBPS/Text/chapter-5.xhtml)
- 结论依据:摘录中关于潜在编码只是未建模原因替身的定义、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 variations in the handwritten digits due to different writers and motor movements as they wrote. With that in mind, we can view P(X| digit, is-handwritten, Z) as the causal Markov kernel of X. That said, it is important to remember that the representation we learn for Z is a stand-in for latent causes and is not the same as learning the actual latent causes.