No Results!

Preliminaries

VC Theory Reproducing Kernel Hilbert Space (RKHS)Hilbert-Schmidt independence criterion (HSIC)

Methmatic Tools

Centered Kernel Alignment (CKA)Mode Connectivity (MC)Hessian Matrix Heavy Tail Self-regression (HT-SR)

Applications

Taxonomizing local versus global structure in neural network loss landscapes Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data A Three-regime Model of Network Pruning Temperature Balancing, Layer-wise Weight Analysis, and Neural Network Training When are ensembles really effective

Home WikiLoss Landscape

Updated on: 2024-05-15

Evaluating natural language processing models with generalization metrics that do not need access to any training or testing data

Prev

Taxonomizing local versus global structure in neural network loss landscapes

Next

A Three-regime Model of Network Pruning

本站由 LinRKen 使用 Stellar 1.27.0 主题创建。
本博客所有文章除特别声明外，均采用 CC BY-NC-SA 4.0 许可协议，转载请注明出处。