Radon Measure Representations for Infinite-Width Neural Networks with Singular Activations
Abstract
The theoretical foundation of infinite-width shallow neural networks relies heavily on continuous integral representations and Barron spaces. Recently, harmonic analysis-specifically the Radon and Ridgelet transforms-has emerged as a powerful tool to invert these representations and compute the optimal network weights. However, a major analytical bottleneck remains: standard neural network activation functions exhibit severe spectral singularities at the frequency origin. To bypass this divergence, existing frameworks either restrict the theory to specific activation families or mathematically quotient out the network's affine components, which inherently limits their practical scope.In this paper, we overcome these limitations by introducing a purely distributional framework for the generalized Radon transform R σ that operates on a broad class of tempered distribution activation functions. By defining a regularized spectrum formulation g(ρ) = (iρ) α σ(ρ), we rigorously absorb the origin singularities without truncating the underlying functional space. Building upon this exact reconstruction, we show that under a definite parity assumption on the activation, there exists an exact linear isometry between Barron functions and their optimal weight measures.
Related
- Fully Deterministic Shallow Neural Networks with Localized Activations. Part II: Condition Number Estimates
- Fully Deterministic Shallow Neural Networks with Localized Activations. Part I: Construction and Optimal Sobolev Approximation
- Sharp Approximation on the Sphere by Linearized Shallow Networks with Analytic Activations
- Best proximity point results of operator type condensing mappings via measure of noncompactness and application to an infinite system of differential equations in c0 and ℓ1
- The singular Gaussian scaling limit for a tagged particle in the infinite-dimensional Dyson model at β=2
BibTeX
@misc{dus2026radon,
title = {{Radon Measure Representations for Infinite-Width Neural Networks with Singular Activations}},
author = {Mathias Dus},
howpublished = {HAL},
year = {2026},
month = jul,
doi = {10.48550/arxiv.2607.29258},
eprint = {2607.29258},
archivePrefix = {arXiv},
url = {https://hal.science/hal-05708168v3},
note = {CC BY-NC-ND 4.0},
}
RIS
TY - UNPB
TI - Radon Measure Representations for Infinite-Width Neural Networks with Singular Activations
AU - Mathias Dus
PY - 2026
DA - 2026/07/31
PB - HAL
DO - 10.48550/arxiv.2607.29258
UR - https://hal.science/hal-05708168v3
AB - The theoretical foundation of infinite-width shallow neural networks relies heavily on continuous integral representations and Barron spaces. Recently, harmonic analysis-specifically the Radon and Ridgelet transforms-has emerged as a powerful tool to invert these representations and compute the optimal network weights. However, a major analytical bottleneck remains: standard neural network activation functions exhibit severe spectral singularities at the frequency origin. To bypass this divergence, existing frameworks either restrict the theory to specific activation families or mathematically quotient out the network's affine components, which inherently limits their practical scope.In this paper, we overcome these limitations by introducing a purely distributional framework for the generalized Radon transform R σ that operates on a broad class of tempered distribution activation functions. By defining a regularized spectrum formulation g(ρ) = (iρ) α σ(ρ), we rigorously absorb the origin singularities without truncating the underlying functional space. Building upon this exact reconstruction, we show that under a definite parity assumption on the activation, there exists an exact linear isometry between Barron functions and their optimal weight measures.
ER -