Datasets:

MInAlA
/

crosscoder-multilayer-split-activations

Tags:

mechanistic-interpretability

sparse-autoencoder

License:

Dataset card Files Files and versions

crosscoder-multilayer-split-activations / v2-differentlayer /smollm3-grpo /singlelayer_separate /base_L19__aligned_L17 /plots

1 contributor

History: 1 commit

ishangarg183's picture

Copy assembled_activations to v2-differentlayer (batch 9/12)

6364775 verified 11 days ago

cf_distribution_per_class.png

174 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
cf_shift_per_class.png

127 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
class_distribution.png

166 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
loss_curves.png

342 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
rho_histogram.png

84.5 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
rho_theta_scatter.png

349 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
shared_geometry_histograms.png

148 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
shared_geometry_theta_vs_norm_ratio.png

425 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago
superposition_analysis.png

192 kB
xet

Copy assembled_activations to v2-differentlayer (batch 9/12) 11 days ago