Each block's code geometry over x, from the covariance of its codes on the inputs where it is active: stable rank ‖C‖²_F / ‖C‖²_2, participation ratio (tr C)² / ‖C‖²_F and effective rank (the exponential of the eigenvalues' entropy). The paper finds a block's dimension saturates at the concept's (2 to 4 for DINOv3) however large b is.
More...
#include <block_sparse_featurizer.hpp>
Each block's code geometry over x, from the covariance of its codes on the inputs where it is active: stable rank ‖C‖²_F / ‖C‖²_2, participation ratio (tr C)² / ‖C‖²_F and effective rank (the exponential of the eigenvalues' entropy). The paper finds a block's dimension saturates at the concept's (2 to 4 for DINOv3) however large b is.
◆ effective_rank
| std::vector<double> pulsatrix::BlockGeometry::effective_rank |
◆ firing_rate
| std::vector<double> pulsatrix::BlockGeometry::firing_rate |
◆ mean_effective_rank
| double pulsatrix::BlockGeometry::mean_effective_rank = 0 |
◆ mean_participation_ratio
| double pulsatrix::BlockGeometry::mean_participation_ratio = 0 |
◆ mean_stable_rank
| double pulsatrix::BlockGeometry::mean_stable_rank = 0 |
Means over blocks that fired on at least 2 inputs, weighted by firing rate.
◆ participation_ratio
| std::vector<double> pulsatrix::BlockGeometry::participation_ratio |
◆ stable_rank
| std::vector<double> pulsatrix::BlockGeometry::stable_rank |
per block; 0 where it fired on fewer than 2 inputs
The documentation for this struct was generated from the following file: