Proposes pointwise Riemannian Dimension from feature eigenvalues to derive tighter, representation-aware generalization bounds for deep networks in the nonlinear regime.
Title resolution pending
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
verdicts
UNVERDICTED 4representative citing papers
The space of rank-r core covariances forms a smooth manifold except on a measure-zero set, enabling a partial-isotropy shrinkage estimator for matrix-variate data.
A combinatorial formula is given for the Euler characteristic of the Grassmannian with d hyperplane sections removed, with focus on generic cases, Schubert divisors, and both complex and real settings.
Studies determinantal varieties and ideals of relations for symmetric matrices with zero diagonal blocks arising as Gram matrices in conformal field theory.
citing papers explorer
-
Pointwise Generalization in Deep Neural Networks
Proposes pointwise Riemannian Dimension from feature eigenvalues to derive tighter, representation-aware generalization bounds for deep networks in the nonlinear regime.
-
Covariance Estimation for Matrix-variate Data via Fixed-rank Core Covariance Geometry
The space of rank-r core covariances forms a smooth manifold except on a measure-zero set, enabling a partial-isotropy shrinkage estimator for matrix-variate data.
-
Hyperplane Arrangements in the Grassmannian
A combinatorial formula is given for the Euler characteristic of the Grassmannian with d hyperplane sections removed, with focus on generic cases, Schubert divisors, and both complex and real settings.
-
Gram Matrices for Isotropic Vectors
Studies determinantal varieties and ideals of relations for symmetric matrices with zero diagonal blocks arising as Gram matrices in conformal field theory.