pulsatrix 1.1.0
Loading...
Searching...
No Matches
pulsatrix::SvccaOptions Struct Reference

Svcca()'s settings. More...

#include <representation_similarity.hpp>

Public Attributes

double variance_kept = 0.99
 Keep each representation's fewest top singular directions whose share of the variance exceeds this, as scikit-learn's PCA(n_components=0.99) chooses. In (0, 1].
 
int64_t directions = 0
 If above 0, keep exactly this many directions of each instead (capped at its rank), as Raghu et al. do to compare layers of different widths on equal terms.
 

Detailed Description

Svcca()'s settings.

Member Data Documentation

◆ directions

int64_t pulsatrix::SvccaOptions::directions = 0

If above 0, keep exactly this many directions of each instead (capped at its rank), as Raghu et al. do to compare layers of different widths on equal terms.

◆ variance_kept

double pulsatrix::SvccaOptions::variance_kept = 0.99

Keep each representation's fewest top singular directions whose share of the variance exceeds this, as scikit-learn's PCA(n_components=0.99) chooses. In (0, 1].


The documentation for this struct was generated from the following file: