The Setting
For a natural number , a spectral exponent with , and a bounded interval , consider an orthonormal -tuple of functions in and positive threshold parameters . Testing the finite-interval action on along lower triangular linear combinations of asks whether an upper bound on this action implies the same upper bound on the spectral sum .
The passage from the finite-interval action to the spectral sum runs through paths of real symmetric matrices connecting the diagonal matrix with entries to its negative under a regularized matrix field with regularization parameter and penalty parameter . One tracks the field primitive along a penalized flow of symmetric matrices for small and then passes to the regularization limit .
Matrix Field and Action
For a natural number , the symmetric matrix space is the space of real symmetric matrices✦, and the lower triangular matrix space is the space of real matrices with whenever .✦
The diagonal symmetric matrix of a vector () is the diagonal matrix with -th diagonal entry .✦ For and , the symmetric outer product is the real symmetric matrix with -entry .✦
Given a real parameter , the field normalization constant is the real number defined by (1): (1)✦
With real parameters , , and , and writing , the regularized scalar primitive is the real number defined by (2): (2)✦
For , , an real matrix , and , the matrix pencil is the real matrix defined by (3), where is the identity matrix: (3)✦
At cutoff , for , , , and an real matrix , the truncated matrix field (suppressing , , and ) is the real matrix given by (4), with evaluated via the continuous functional calculus: (4)✦
The regularized matrix field of an real matrix , for , , and , is the real matrix given by the cutoff limit (5): (5) When and , the symmetric matrix field (for and ) is the map sending each to .✦
The matrix field primitive (written when , , and are fixed) of an real matrix , for , , and , is the real number defined by (6): (6)✦
For vectors () and a real parameter , the action density (suppressing and ) is the real number defined by (7): (7)✦ For a bounded interval and a positive diagonal matrix , the finite-interval action of a path is the real number .
On a pair with , , and , the action defect is the real number given by (8): (8)✦
The constraint defect of a pair , for , , and , is the real number defined by the squared Hilbert-Schmidt norm (9): (9)✦
The Claim
Theorem 1. (1) For , real numbers , , , and , a vector , paths and , and such that each component has derivative at , has derivative at in , and , the differential identity (10) holds at : (10)✦
(2) For , real parameters , vectors , and , the pointwise bound (11) holds: (11)✦
(3) If is a topological space, is compact, are continuous functions with for all and whenever satisfies , and , then there exists such that every and satisfy (12): (12)✦
(4) Given , , , , a compact set , and , there exists such that every and satisfy (13): (13)✦
(5) Suppose , , , , , , and . Let be continuous on with such that each component has derivative at every , and let satisfy and , with derivative within in equal to at every . If for all , then (14) holds: (14)✦
(6) Given , , , , , and with for all , suppose are continuous on with , each component has derivative at every , and equals when and otherwise. If for every , then (15) holds: (15)✦
(7) Given , , , , and with for all , suppose are continuous on with , each component has derivative at every , and equals when and otherwise. If for every , then (16) holds: (16)✦
(8) For , , and with for all , the endpoint jump satisfies the right-hand limit (17): (17)✦
In the paper, the result reads: along the path , , yielding with , , and , while compactness of the trajectory set ensures as and dominated convergence as yields the action inequality.
What Is Checked
The claims are stated over natural numbers , real parameters , , , , , , and , the spaces and , the constructions and , the matrix fields and built from , , and the pencil , the primitive , the action density together with and , paths , , , and on satisfying boundary and differential hypotheses (and orthonormality for ), and continuous real functions and on a compact set in a topological space or in .