Metraipolis › archive › dc_6a12d2c10ffa

Neural Approximation by Function Composition: Rigidity and Doubly Exponential Convergence — arXiv:2609.25874

Source
arXiv (en)
Original
https://arxiv.org/abs/2609.25874
License (as archived)
CC0-1.0
Archived
2026-10-08T15:07:36.554Z
SHA-256 of the archived text
32b6a34d001f4375edce83c2bf3b6493f337710afe595685ccc978aa47a4632b
Stable URL of this page
https://metraipolis.com/documento/dc_6a12d2c10ffa

How to cite

Neural Approximation by Function Composition: Rigidity and Doubly Exponential Convergence — arXiv:2609.25874 — arXiv (en), https://arxiv.org/abs/2609.25874 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_6a12d2c10ffa (Metraipolis archive dc_6a12d2c10ffa, sha256 32b6a34d001f4375edce83c2bf3b6493f337710afe595685ccc978aa47a4632b)
BibTeX
@misc{dc6a12d2c10ffa,
  title = {Neural Approximation by Function Composition: Rigidity and Doubly Exponential Convergence — arXiv:2609.25874},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.25874},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_6a12d2c10ffa; sha256 32b6a34d001f4375edce83c2bf3b6493f337710afe595685ccc978aa47a4632b},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.25874. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Deep neural networks approximate functions by composing affine maps with nonlinear activations, but how composition itself creates approximation power is not yet fully understood. We investigate a fundamental mechanism: geometrically weighted sums of iterates of a single scalar generator function. This mechanism underpins the classical tent-map construction of the function \(x - x^2\) and related recursive representations used by Yarotsky, W. E, et al., to analyze the approximation powers of deep neural networks. First, we establish a rigidity theorem: for continuous piecewise linear generators with a finite number of segments, any \(C^3\) function that can be represented in this way is at most quadratic. For non-affine quadratic functions, the geometric factor is at least $1/4$. This result both reveals limitations of the tent-map approach and complements existing methods based on hierarchical bases and recursive polynomial constructions. Second, using an exact remainder identity as guidance, we construct a smooth generator whose iterates yield doubly exponential error decay in total depth for square approximation and, through multiplication modules, for each fixed polynomial. For power series with absolutely summable coefficients on \([-1,1]^d\), distributing depth according to monomial degree yields a uniform approximation error of order \(O(e^{-cL^{1/d}})\) on each interior cube. These findings demonstrate how generator dynamics and remainder estimates govern depth allocation and approximation rates of deep neural networks.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-23 in cs.LG).
Categorie arXiv: Machine Learning (cs.LG) ; Information Theory (cs.IT).
Pagina del lavoro: https://arxiv.org/abs/2609.25874 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-08T14:37:29.437Z dalla pagina https://arxiv.org/catchup/cs.LG/2026-09-23?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).