The Setting
Take a finite set , a mesh step , and a triple of complex matrices. Under Gaussian heat evolution on the zero-sum plane of three center coordinates, a matrix energy functional dissipates along scales through a Hessian quadratic cost, while output-dependent cutoff endpoints retain each entry only across the half-open scale window .
The problem is to bound both the total energy jumps produced by bounded-variation replacements of and the scale integral of the masked Hessian cost in terms of a cubic norm of , uniformly in , , the cutoff matrices and , and the number of distinct endpoint values.
Triangle Matrices and Matrix Energy
The triangle vertex set is the additive group under cyclic addition. For an index set , the triangle block space is the Cartesian product of the triangle vertex set and . For a complex square matrix on , the real trace is the real part of the trace of . For a complex square matrix on , the matrix energy is the real number , where is the functional-calculus square root of . For complex square matrices and on , the real Hilbert-Schmidt inner product is the real number .
Given a complex square matrix on , the Sylvester operator is the linear operator on complex square matrices on defined on each complex square matrix on by . Given complex square matrices and on , the Sylvester penalty is the real number . For complex square matrices and on , the Sylvester variational functional is the real functional defined on each complex square matrix on by . For complex square matrices and on , the root Hessian cost is the value . For complex square matrices and on , the Hessian cost is the root Hessian cost evaluated at and .
For a matrix indexed by whose entries are complex matrices, the block matrix assembly of is the complex matrix on whose entry is . For a family of complex matrices indexed by , the triangle matrix blocks form the block matrix whose block is when , when , and when , and the triangle matrix is the block matrix assembly of . For , the vertex block projection on is the diagonal matrix with entry at when and otherwise. For a family of complex matrices indexed by and , the vertex star row is the matrix . The vertex star Gram matrix of a family of complex matrices indexed by and is the matrix .
For a family of complex matrices indexed by and a complex matrix , the zero-component replacement is the matrix family indexed by equal to at and to at . Given a matrix family on and real node weights on indexed by , the weighted matrix array is the matrix family on whose entry at vertex is . For real node weights on indexed by , the vertex root matrix on is the diagonal matrix with entry at . Given a matrix family on , real node weights on , and , the vertex star base row is the matrix . The vertex star base column of a matrix family on , real node weights on , and is the matrix . For a complex square matrix on an index set and a real weight function , the weighted Gram matrix of with weight is the matrix .
Gaussian Heat Costs and Masked Truncations
For real variance and , the Gaussian density is (1) Given step size , point , center , and variance , the heat weight is the real number . Given point , center , and variance , the Gaussian score is the real number . For a step size , node coordinates , centers , and variances , the vertex heat weights form the family of node weights on defined at and by . Given a matrix family on indexed by and parameters , the heat matrix array is the weighted matrix array formed with the vertex heat weights for . For a vertex , the heat matrix energy is the real number .
The fixed base matrix of a matrix family on indexed by , real node weights on , and vertices is the vertex star base row when , and the vertex star base column when . Given , coordinates , a group assignment , centers , and variances , the grouped heat weight at is the real number . For group , the grouped first variation weight at is the real number equal to if and otherwise. For a matrix family on indexed by , , , , and , let denote the vertex heat weights for , and set , , , and : then the local Gram matrix is the matrix given by the weighted Gram matrix of with the grouped heat weight, and the local first variation matrix is the matrix given by the weighted Gram matrix of with the grouped first variation weight for group . The heat Hessian cost of at is the real number .
At vertex and sum level , the plane coordinate chart is the map sending each to the triple summing to equal to at vertex and to at the remaining two vertices in increasing order. For a function and , the plane integral over the constant-sum center plane is (2) For a matrix family on indexed by , step size , node coordinates , sum level , and variances , the total plane heat energy is (3)
Given scale endpoints , scale , and , the window mask is the complex number equal to when and otherwise.✦ Given a matrix family on indexed by , real cutoff matrices , and scale , the masked array is the zero-component replacement with given by . For a vertex , the strict diffusion rate is the function defined by for and for . For a matrix family on indexed by , real cutoff matrices , step size , node coordinates , vertex , and scale , the masked heat cost is (4) For a matrix family on indexed by and step size , the cubic array norm is (5)
Discrete Maximal Functions
The discrete interval of radius centered at an integer is the set . The support radius of a finite set around is the supremum , taken to be when is empty.✦ Given a finite set , a function , , and , the discrete interval average is (6) and the discrete maximal function of over at is .
Given a finite set and a function , the extension by zero is the function equal to for and for .✦ For a finite set , a function , and , the line maximal function is the real number . Given a finite set , a family of complex matrices indexed by , and indices , the jump maximal function is (7)
Dissipation and Jump Constants
The strict positivity constant is the real number (8)✦ The mesh mass constant is the real number .✦ The jump scalar constant is the real number .✦ The masked constant is the real number .✦
The Claim
Theorem 1. Let be a finite set and be a family of complex matrices indexed by . Then the jump maximal sum satisfies (9)✦ Suppose further that , and define the uniform grid by for . For any real cutoff matrices with for all , any nonempty finite set containing every and with for all , and any , the masked heat cost satisfies (10)✦ Let , and let be a matrix sequence satisfying for all and and for all . For any variance profiles with for all and , the total plane energy jumps satisfy (11)✦ For any nondecreasing sequence with and any , the integrated masked dissipation satisfies (12)✦
In the paper, the result reads: for every , integer , complex arrays on , and endpoints , taking any permutation of and gives with a constant independent of the arrays, mesh, endpoints, and number of distinct endpoint values, where the diagonal heat Hessian cost at stands for the paper’s and are the extremal scale endpoints. The library states this bound on an arbitrary finite set rather than , uses general half-open cutoffs in a finite set with integration limits and in (10), parametrizes the permutations by for with , and states separately the jump maximal bound (9), the total energy jump bound (11), and the stepwise masked dissipation bound (12).
What Is Checked
The jump maximal bound (9) is stated over any finite set and complex matrix family ; over the same and with mesh step and uniform grid , the masked cost integral bound (10) is stated over real cutoff matrices in a nonempty finite set and , while the total energy jump bound (11) and finite masked dissipation bound (12) are stated over any and matrix sequence satisfying and , together with variance profiles in (11) or a nondecreasing scale sequence with and in (12).