gradient-analysis

Here are 2 public repositories matching this topic...

bharatmahaur / LeNo

Code for paper: "Improved Residual Network Based on Norm-Preservation for Visual Recognition" https://doi.org/10.1016/j.neunet.2022.10.023

deep-neural-networks computer-vision deep-learning pytorch imagenet image-recognition resnet convolutional-neural-networks residual-network resnext gradient-flow visual-recognition wideresnet optimization-problem gradient-analysis norm-preservation gradient-norm leno-resnet optimization-stability

Updated Oct 31, 2022
Python

royxforge / loss-landscape-analysis

Star

Empirical comparison of MSE vs Cross-Entropy on MNIST. CE reaches 97% accuracy in 4 epochs vs 13 for MSE - gradient norm analysis confirms the saturation mechanism.

deep-learning pytorch mnist classification mlp loss-functions cross-entropy mean-squared-error gradient-analysis gradient-saturation

Updated Jul 20, 2026
Python

Improve this page

Add a description, image, and links to the gradient-analysis topic page so that developers can more easily learn about it.

Curate this topic

Add this topic to your repo

To associate your repository with the gradient-analysis topic, visit your repo's landing page and select "manage topics."

Learn more