SASA replaces single-vector decoders in SAEs with learned subspaces plus block sparsity and nuclear-norm regularization, proving that a single group becomes the global minimizer once block size meets intrinsic dimension and yielding polynomial rather than exponential sample complexity.
Title resolution pending
4 Pith papers cite this work, alongside 455 external citations. Polarity classification is still indexing.
4
Pith papers citing it
455
external citations · OpenAlex
citation-role summary
method 1
citation-polarity summary
years
2026 4roles
method 1polarities
use method 1representative citing papers
Ultrametric graphons model hierarchical community networks and yield closed-form Laplacian spectra that approximate those of sampled random graphs with high probability as hierarchy depth grows.
Anchor PCA recovers a maximal invariant subspace for multi-domain data via PCA on a modified target matrix that trades off explained variance with domain agreement.
A PC-based decomposition of FVE into low- and high-dimensional components reduces bias when applying GWASH or LMM-REML to strongly correlated high-dimensional predictors.
citing papers explorer
-
Subspace-Aware Sparse Autoencoders for Effective Mechanistic Interpretability
SASA replaces single-vector decoders in SAEs with learned subspaces plus block sparsity and nuclear-norm regularization, proving that a single group becomes the global minimizer once block size meets intrinsic dimension and yielding polynomial rather than exponential sample complexity.
-
Ultrametric Graphons and Hierarchical Community Networks: Spectral Theory and Applications
Ultrametric graphons model hierarchical community networks and yield closed-form Laplacian spectra that approximate those of sampled random graphs with high probability as hierarchy depth grows.
-
Anchor PCA
Anchor PCA recovers a maximal invariant subspace for multi-domain data via PCA on a modified target matrix that trades off explained variance with domain agreement.
-
Principal Components Decomposition of Fraction of Variance Explained in High Dimensional Linear Models with Strong Correlation
A PC-based decomposition of FVE into low- and high-dimensional components reduces bias when applying GWASH or LMM-REML to strongly correlated high-dimensional predictors.