原书出处
Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)
relu layer maps many points to the axes of the positive quadrant. In general, with relu-nets, a lot of data density will build up along these axes, because any point not in the positive quadrant snaps to an axis. This effect becomes exaggerated in real networks with high-dimensional embeddings. In particular, for a width-N layer, the region that is strictly positive occupies only proportion of
原书定位详情
来源:a19c7933-4eda-4176-aebd-30fe7e1ad292
正文位置:extracted/Foundations of Computer Vision (Antonio Torralba, Phillip Isola etc.)/_epub-src/OPS/xhtml/c013.xhtml
EPUB SHA-256:60a7eebc43f5d9a675d776fed3736fbdaa791e55877023f2f8cba016a9b384f8