arXiv ScienceSearch

arXiv · 2506.22693

Proof-Carrying Analytic Approximation: Local-to-Global Evidence Transport at Encoding Cost

Abstract

Under quasi-uniform refinement, bounded local-encoding hypotheses, and local $W^{r,2}$ approximation of order $r\ge 2$ in a rational piecewise-polynomial presentation of $W^{1,2}(0,1)$, carrying the complete proof genealogy up to the level required by an accuracy $\varepsilon$ costs the same asymptotic bit order as the finest-level conventional coefficient encoding. If $B_n=Θ(M_nβ_n)$ denotes that level-$n$ encoding size, our compiler transports supplied local approximation and overlap witnesses through exact partition-of-unity synthesis and geometric refinement to a represented limit with total certificate size $O(B_{m(\varepsilon)})$, where $m(\varepsilon)=O(\log(1/\varepsilon)/(r-1))$. The construction makes no oracle query to an independently supplied semantic target name ($Q_{\rm target}=0$). When $β_n=O(n+1)$, this becomes $O(\varepsilon^{-1/(r-1)}(1+\log(1/\varepsilon)))$. The surrounding framework is intentionally separated from this resource theorem. Every real computable Banach presentation admits a uniformly computable linear isometric embedding into standard computable $C([0,1])$, with computable inverse on its represented range. Complete metric evidence with rational strict slack collapses extensionally to the represented analytic metric once effective names are available, while chosen evidence transformations retain construction history and resource information. For the Lipschitz grammar used here, qualitative evidence-local lifting is canonical; the nontrivial question is therefore which evidence is retained and at what cost.

Explore related subjects

Keep this discovery

BibTeXRIS

Andreu Ballus Santacana. 2026-08-31. Proof-Carrying Analytic Approximation: Local-to-Global Evidence Transport at Encoding Cost. https://arxiv.org/abs/2506.22693

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related papers

A categorical formulation of Kraus' paradox

We give a categorical formulation of Kraus' "magic trick" for recovering information from truncated types. Rather than type theory, we work in Van den Berg-Moerdijk path categories with a univalent universe, and rather than propositional truncation we work with arbitrary cofibrations, which includes truncation as a special case. We show, using Kraus' argument that any cofibration with homogeneous domain is a monomorphism. We give some simple concrete examples in groupoids to illustrate the interaction between homogeneous types, cofibrations and univalent fibrations.

math.CT

Further Comments on Yablo's Construction

We continue our analysis of Yablo's coding of the liar paradox by infinite acyclic graphs. The present notes are based on and continue the author's previous results on the problem. In particular, our approach is often more systematic than before.

math.CO

Quantified propositional calculi and narrow implicit proofs

In the implicit version of a propositional proof system Q, we work with Q-proofs that are not written down directly, but are succinctly encoded by circuits. Thus implicit Q-proofs are potentially exponentially shorter than usual Q-proofs. We study narrow implicit proofs, a restricted version of this notion, in which lines in the encoded proof can only have polynomial size. We use a cut-elimination construction to show that G_{i+1} is equivalent to narrow implicit G_i, for i >= 1, where G_i is the extension of Frege allowing reasoning with Sigma^q_i quantified propositional formulas. We show that G_1 is equivalent to implicit resolution.

cs.LO