Algebraic Invariants of Lightning Self-Attention

• URL: http://arxiv.org/abs/2604.15632v1
• Date: Fri, 17 Apr 2026 02:16:49 GMT
• Title: Algebraic Invariants of Lightning Self-Attention
• Abstract summary: We study the coefficients of lightning self-attention as coordinates of an algebraic variety.<n>We identify linear and nonlinear families of algebraic invariants, including Chow-type, low-rank, Veronese-type, and Sylvester-based constraints.
• Score: 11.737898657629017
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: We study the polynomial coefficients of lightning self-attention as coordinates of an algebraic variety. We identify linear and nonlinear families of algebraic invariants, including Chow-type, low-rank, Veronese-type, and Sylvester resultant-based constraints.

Related papers

• The Dynamical Lie Algebra of QAOA-MaxCut on the Complete Graph [9.937940252962443]
  We show that the variance of the associated loss function scales linearly in the number of qubits.<n>This solves an open problem from [ASYZ26] and confirms that such systems do not exhibit barren plateaus.
  arXiv  Detail & Related papers  (2026-07-01T13:45:02Z)
• Concentration of measure for non-linear random matrices with applications to neural networks and non-commutative polynomials [0.0]
  We prove concentration inequalities for several models of non-linear random matrices.<n>We obtain estimates for linear statistics of the conjugate kernel of neural networks and neural non-commutatives in (possibly dependent) random matrices.
  arXiv  Detail & Related papers  (2025-07-10T10:47:42Z)
• Division polynomials for arbitrary isogenies [0.24864093375172572]
  We extend the definition of divisions to arbitrary isogenies of elliptic curves, including those whose kernels do not sum to the identity.<n>In analogy to the case of divisions for multiplication-by-n, we demonstrate relations, identities relating to classical elliptic functions.
  arXiv  Detail & Related papers  (2025-03-19T17:10:31Z)
• Enriching Diagrams with Algebraic Operations [49.1574468325115]
  We extend diagrammatic reasoning in monoidal categories with algebraic operations and equations. We show how this construction can be used for diagrammatic reasoning of noise in quantum systems.
  arXiv  Detail & Related papers  (2023-10-17T14:12:39Z)
• Dynamical chaos in nonlinear Schr\"odinger models with subquadratic power nonlinearity [137.6408511310322]
  We deal with a class of nonlinear Schr"odinger lattices with random potential and subquadratic power nonlinearity. We show that the spreading process is subdiffusive and has complex microscopic organization. The limit of quadratic power nonlinearity is also discussed and shown to result in a delocalization border.
  arXiv  Detail & Related papers  (2023-01-20T16:45:36Z)
• Supersymmetry and Shape Invariance of exceptional orthogonal polynomials [0.0]
  We express the differential equations for the Jacobi and the Laguerre exceptional orthogonals (EOP) as the eigenvalue equations. We show that the underlying shape invariance is responsible for the solubility of the differential equations associated with these systems.
  arXiv  Detail & Related papers  (2022-06-08T13:54:42Z)
• Polynomial algebras of superintegrable systems separating in Cartesian coordinates from higher order ladder operators [0.618778092044887]
  We introduce the general algebras characterizing a class of higher order superintegrable systems that separate in coordinates. The construction relies on underlying Heisenberg algebras and their defining higher order ladder operators.
  arXiv  Detail & Related papers  (2022-02-27T03:33:26Z)
• Convolutional Filtering and Neural Networks with Non Commutative Algebras [153.20329791008095]
  We study the generalization of non commutative convolutional neural networks. We show that non commutative convolutional architectures can be stable to deformations on the space of operators.
  arXiv  Detail & Related papers  (2021-08-23T04:22:58Z)
• Hessian Eigenspectra of More Realistic Nonlinear Models [73.31363313577941]
  We make a emphprecise characterization of the Hessian eigenspectra for a broad family of nonlinear models. Our analysis takes a step forward to identify the origin of many striking features observed in more complex machine learning models.
  arXiv  Detail & Related papers  (2021-03-02T06:59:52Z)
• Some Hoeffding- and Bernstein-type Concentration Inequalities [47.24550702683417]
  We prove concentration inequalities for functions of independent random variables under sub-gaussian and sub-exponential conditions. The utility of the inequalities is demonstrated by an extension of the now classical method of Rademacher complexities to Lipschitz function classes and unbounded sub-exponential distribution.
  arXiv  Detail & Related papers  (2021-02-11T23:09:13Z)
• Algebraic Neural Networks: Stability to Deformations [179.55535781816343]
  We study algebraic neural networks (AlgNNs) with commutative algebras. AlgNNs unify diverse architectures such as Euclidean convolutional neural networks, graph neural networks, and group neural networks.
  arXiv  Detail & Related papers  (2020-09-03T03:41:38Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.