2581 valuable arXiv mathematics papers the pipeline has analyzed — each with a short note on what its math gives to machine learning.
The paper gives a constructive analysis of how different two-point splittings of a quadratic nonlinearity cancel or amplify collocation aliasing, and shows that the splitting minimizing instantaneous integration error need not be the one…
The paper gives an exact finite generating system for smooth vector fields whose coefficients vanish to a prescribed order at a designated point. This can be transferred into neural dynamical models by factoring the vector field through…
The paper provides a slope-dependent anisotropic diffusion law for vector-valued graph outputs. Diffusion is controlled by the induced graph metric, so the layer smooths curvature without applying isotropic blur indiscriminately. The most…
The paper gives a constructive, 1-Lipschitz embedding of persistence diagrams and barcodes equipped with Wasserstein-type matching distances into functions on \(\mathbb{N}\times\mathbb{R}\). The transferable asset is that a…
The paper provides a nonstandard structural result: under a semiglobal Polyak–Łojasiewicz inequality, a smooth objective can still be reduced to nonlinear least squares even when the usual global PŁI fails at large loss values. The key…
The paper develops an exterior differential calculus on supports of Radon measures that may be discrete, singular, or otherwise non-manifold, while retaining restriction, extension, pullback, cohomology, and Stokes-type constructions. The…
The paper provides explicit global bounds on Kemeny's constant in terms of graph diameter, resistance distances, and spectral gaps. This suggests a practical regularizer for learned graph structures in GNNs: discourage propagation graphs…
The paper supplies an explicit graded, noncommutative state law for first-order increments together with antisymmetric second-order area and symmetric quadratic variation. The transferable asset is that sequential composition is exact…
The paper offers a transferable multi-scale similarity construction: instead of choosing one kernel bandwidth, it selects several bandwidths from empirical quantiles of pairwise squared distances. This is useful for neural representation…
The paper supplies an explicit scale-consistent coupling between low-order fractional variation, an Lp amplitude norm, and high-order fractional energy through a sharp homogeneous Gagliardo–Nirenberg inequality. The transferable asset is…
The paper constructs a nontrivial Markov kernel on length-n empirical count vectors by taking the stationary one-step conditional law of overlapping windows, even though the count process itself is not Markov. Its key transferable asset is…
The paper provides a distortion-aware variational template rather than merely a mesh-processing algorithm: it jointly minimizes local anisotropy and nonuniform pushed-forward mass while allowing the coordinate domain itself to adapt. This…
The paper gives a constructive way to represent and optimize an exactly doubly stochastic nonnegative matrix without materializing its dense entries. Its key asset is the shared latent marginal: two independent row-simplex factors U and V…
The paper provides a constructive mechanism by which hard clipping creates memory: the clipped-off latent excess is stored and partially injected into the next time step, even when the external shocks are independent. Its transferable…
The paper's transferable contribution is a deterministic ambient-domain representation that lets one spectral operator handle samples defined on different physical geometries and discretizations. Each geometry is encoded by a…
The paper gives a constructive tree recursion for a random-cluster model with q<1, where the nontrivial fixed point appears exactly above p_c=q/(\Delta+q-2), together with conditional negative association of connectivity indicators on…
The paper's transferable asset is a solver-level differentiation rule: when a forward computation is an ordered sequence of local implicit block solves, its exact discrete adjoint is obtained by reverse-order local adjoint solves rather…
The paper offers a transferable phase-transition mechanism: Lee-Yang zeros of the complexified partition function approach the real thermodynamic axis at a physical transition, so the partition-function modulus near that axis is a…
The paper exposes a useful equivalence between a pairwise inner-product cut objective and vector balancing: maximizing cross-part interactions is equivalent, up to a constant, to minimizing the squared norms of aggregate feature vectors…
The paper identifies a sharp spectral signature of near-equality in the Gaussian (B)-theorem: the covariance eigenvalues of the Gaussian restricted to a symmetric convex body are polarized, with each principal variance close either to 0 or…
The paper's transferable asset is an anchored fixed-point principle for monotone variational inequalities: rather than repeatedly applying an unanchored high-order solver, it blends each approximate solution with a fixed initial anchor…
The paper gives a constructive binary edge labeling for trees such that adjacent vertices have different weighted incident-edge sums. This can transfer to tree-structured GNNs as a deterministic symmetry-breaking positional encoding…
The paper gives a sharp minimax law for progressively refining a conserved positive mass into binary pieces: no binary refinement schedule can keep the largest-to-smallest piece ratio below D_n = 2^{1-1/\lceil n/2\rceil} at every…
The paper develops lexicographic functional calculus, an ordered noncommutative calculus for differentiating matrix functions with perturbation operators inserted between spectral factors. Its transferable asset is an explicit higher-order…