Modules

Custom PyTorch modules.

Orthogonalization layers.

• OrthogonalLayer - Base class for orthogonalization layers
• PCALayer - PCA-based orthogonalization
• QRLayer - QR decomposition orthogonalization
• RunningQRLayer - Running QR orthogonalization
• SVDLayer - SVD-based orthogonalization

Standardization and normalization.

• StandardizeLayer - Standardization layer
• WeightedBatchNorm1d - Weighted batch normalization

Activation functions.

• ShiftedSoftplus - Shifted softplus activation
• SmoothedShiftedSoftplus - Smoothed shifted softplus
• TemperedSigmoid - Tempered sigmoid activation
• TemperedSine - Tempered sine activation