Riemannian optimization, alternating truncated-SVD/Procrustes steps, and binary-search rank selection are presented as a framework for finding orthogonal disentangler rotations that reduce tensor-network bond dimensions.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
math.NA 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Numerical Optimization for Tensor Disentanglement
Riemannian optimization, alternating truncated-SVD/Procrustes steps, and binary-search rank selection are presented as a framework for finding orthogonal disentangler rotations that reduce tensor-network bond dimensions.