Deep Learning as Ricci Flow

arXiv:2404.14265v1 Announce Type: new
Abstract: Deep neural networks (DNNs) are powerful tools for approximating the distribution of complex data. It is known that data passing through a trained DNN classifier undergoes a series of geometric and topological simplifications. While some progress has been made toward understanding these transformations in neural networks with smooth activation functions, an understanding in the more general setting of non-smooth activation functions, such as the rectified linear unit (ReLU), which tend to perform better, is required. …

abstract arxiv classifier cs.lg data deep learning distribution dnn flow functions math.dg networks neural networks progress series through tools type understanding