all AI news
A Study of Neural Collapse Phenomenon: Grassmannian Frame, Symmetry and Generalization. (arXiv:2304.08914v2 [cs.LG] UPDATED)
cs.LG updates on arXiv.org arxiv.org
In this paper, we extend original Neural Collapse Phenomenon by proving
Generalized Neural Collapse hypothesis. We obtain Grassmannian Frame structure
from the optimization and generalization of classification. This structure
maximally separates features of every two classes on a sphere and does not
require a larger feature dimension than the number of classes. Out of curiosity
about the symmetry of Grassmannian Frame, we conduct experiments to explore if
models with different Grassmannian Frames have different performance. As a
result, we discover …
arxiv classification feature features generalized hypothesis neural collapse optimization paper sphere study symmetry