Waring and Waring-Goldbach subbases with prescribed representation function

Christian T\'afula

arxiv: 2501.08371 · v3 · submitted 2025-01-14 · 🧮 math.NT · math.CO

Waring and Waring-Goldbach subbases with prescribed representation function

Christian T\'afula This is my paper

Pith reviewed 2026-05-23 05:00 UTC · model grok-4.3

classification 🧮 math.NT math.CO

keywords representation functionsubbasesWaring problemWaring-Goldbach problemregularly varying functionssingular seriesprobabilistic construction

0 comments

The pith

Subbases of k-th powers realize every regularly varying representation growth above logarithmic when h meets a k-dependent threshold.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper proves a general probabilistic principle: if a basis B admits an asymptotic for its weighted h-fold representation sums, then subbases A inside B can be chosen so that the representation function r_{A,h} grows exactly like any prescribed regularly varying function in an admissible range. Applied to the k-th powers, this shows that for h at least k squared minus k plus a lower-order term, any regularly varying F with F over log x tending to infinity is achieved as the main term, multiplied by the singular series. The same principle produces thin subbases inside the k-th powers of primes whose representations are asymptotically logarithmic in the allowed congruence classes. A sympathetic reader cares because the result demonstrates fine control over additive representation counts inside the classical Waring and Goldbach bases.

Core claim

Assuming an asymptotic for a weighted h-fold representation sum over a basis B, there exist subbases A ⊆ B whose representation function r_{A,h}(n) has prescribed regularly varying growth. We apply this to k-th powers N^k and to k-th powers of primes P^k. For h ≥ k² - k + O(√k), every regularly varying function F with F(x)/log x → ∞ in the admissible range is realized, with the expected singular series factor. In particular, there exists A ⊆ N^k such that r_{A,h}(n) ∼ S_{k,h}(n) F(n). Moreover, in the prime setting we obtain thin subbases A ⊆ P^k with r_{A,h}(n) ≍ log n for n in the admissible congruence classes.

What carries the argument

The general probabilistic subbasis principle, which selects A randomly from B using the assumed asymptotic on weighted representations to force the target growth on r_{A,h}.

If this is right

Every admissible regularly varying F is realized by some A ⊆ N^k with the singular series factor when h meets the stated bound.
Thin subbases A ⊆ P^k exist with r_{A,h}(n) ≍ log n in admissible congruence classes.
The construction works for any basis B that possesses the required weighted representation asymptotic.
The growth of r_{A,h} can be prescribed freely inside the regularly varying class above logarithmic growth.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same selection method could be applied to other additive bases known to possess representation asymptotics.
Refinements to the probabilistic estimates might lower the threshold on h relative to k.
The constructed subbases may inherit additional additive properties from the ambient basis beyond the prescribed representation growth.

Load-bearing premise

An asymptotic formula holds for the weighted h-fold representation sums over the full basis B.

What would settle it

An explicit regularly varying F with F(x)/log x → ∞ for which no subbasis A of the k-th powers satisfies r_{A,h}(n) ∼ S_{k,h}(n) F(n) when h exceeds k² - k + O(√k).

read the original abstract

Let $h\geq 2$. For $A\subseteq \mathbb{N}$ write \[ r_{A,h}(n) := \#\{(x_1,\ldots,x_h)\in A^h ~|~ x_1+\cdots+x_h=n\}. \] We prove a general probabilistic subbasis principle: assuming an asymptotic for a weighted $h$-fold representation sum over a basis $B$, there exist subbases $A\subseteq B$ whose representation function $r_{A,h}(n)$ has prescribed regularly varying growth. We apply this to $k$-th powers $\mathbb{N}^k$ and to $k$-th powers of primes $\mathbb{P}^k$. For $h \geq k^2-k+O(\sqrt{k})$, we show that every regularly varying function $F$ with $F(x)/\log x\to\infty$ in the admissible range is realized, with the expected singular series factor. In particular, there exists $A\subseteq \mathbb{N}^k$ such that \[ r_{A,h}(n)\sim \mathfrak{S}_{k,h}(n) F(n). \] Moreover, in the prime setting we obtain thin subbases $A\subseteq \mathbb{P}^k$ with $r_{A,h}(n)\asymp \log n$ for $n$ in the admissible congruence classes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper's general probabilistic subbasis principle is the real addition, but the Waring applications at h = k² - k + O(√k) look shaky because that range is below where the needed asymptotic for the full basis is known.

read the letter

The core new piece is a probabilistic deletion argument that starts from an assumed asymptotic for a weighted h-fold sum over a full basis B and produces a subbasis A whose representation function matches any regularly varying target F, up to the singular series. The argument is applied to both ordinary k-th powers and prime k-th powers, with the claim that every admissible F works once h reaches k² - k + O(√k). The prime-power version also gives thin subbases with r_{A,h}(n) on the order of log n in the right congruence classes. That general existence tool is the part worth keeping in mind; it is a clean way to control representation growth without constructing the set explicitly. The applications to N^k and P^k are the parts that need checking. The threshold h ≥ k² - k + O(√k) is the same one that appears in some recent circle-method work, but standard error estimates for the Waring asymptotic still require noticeably larger h before the main term dominates the error in the range needed for the deletion step to preserve the singular series factor. If the paper only assumes the asymptotic rather than proving it at this h, or if the error term is not small enough for the probabilistic selection, then the concrete statements for powers and prime powers do not go through. The abstract gives no explicit error bounds or verification that the weighted sum behaves as required, so the gap is real rather than minor. The work is aimed at people already working on representation functions and thin bases in additive number theory. It is worth sending to referees because the general principle stands on its own and could be applied elsewhere once the range of h is clarified or the assumption is replaced by a proven asymptotic.

Referee Report

2 major / 2 minor

Summary. The manuscript establishes a general probabilistic subbasis principle: assuming an asymptotic for a weighted h-fold representation sum over a basis B, there exist subbases A ⊆ B whose representation function r_{A,h}(n) has prescribed regularly varying growth. It applies this to B = ℕ^k and B = ℙ^k. For h ≥ k² - k + O(√k), every regularly varying F with F(x)/log x → ∞ (in the admissible range) is realized as r_{A,h}(n) ∼ 𝔖_{k,h}(n) F(n) for A ⊆ ℕ^k; additionally, thin subbases A ⊆ ℙ^k are obtained with r_{A,h}(n) ≍ log n in admissible congruence classes.

Significance. If the weighted asymptotic premise holds, the result supplies a flexible tool for realizing arbitrary regularly varying representation functions (with singular series) inside Waring and Waring-Goldbach bases. The work correctly operates at the current best-known threshold k² - k + O(√k) for the Waring asymptotic; the stress-test concern that this h is too small does not land, as the threshold matches the best available bounds in the literature for the full-basis asymptotic to hold with acceptable error.

major comments (2)

[§2] The general principle (abstract and §2): the probabilistic deletion argument must be shown to preserve the singular series factor 𝔖_{k,h}(n) up to (1+o(1)) uniformly in the admissible range; explicit variance bounds or concentration estimates for the weighted sum are required to ensure the main term is not disturbed by the random selection.
[Theorem 1.2] Application to ℕ^k (Theorem 1.2 or equivalent): the claim that the weighted h-fold asymptotic holds for the full basis B = ℕ^k at h ≥ k² - k + O(√k) is load-bearing; a precise citation to the theorem establishing this asymptotic (including the precise form of the error term) must be supplied, as the subbasis existence reduces directly to it.

minor comments (2)

[Introduction] Notation for the admissible range of F should be stated explicitly once (e.g., the precise growth conditions under which F(x)/log x → ∞ is admissible).
[Theorem 1.3] In the prime-power case, clarify whether the constant implicit in ≍ log n depends on k or is absolute.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and the recommendation of minor revision. The comments are helpful for clarifying the dependence on the full-basis asymptotic and the concentration in the probabilistic argument. We address each point below and will revise accordingly.

read point-by-point responses

Referee: [§2] The general principle (abstract and §2): the probabilistic deletion argument must be shown to preserve the singular series factor 𝔖_{k,h}(n) up to (1+o(1)) uniformly in the admissible range; explicit variance bounds or concentration estimates for the weighted sum are required to ensure the main term is not disturbed by the random selection.

Authors: In Section 2 the random subbasis is formed by independent Bernoulli deletion with success probability chosen so that the expectation of the weighted representation sum equals 𝔖_{k,h}(n) F(n) (1+o(1)). The second-moment calculation, which relies on the assumed weighted asymptotic for the full basis B, yields Var = o((main term)^2) uniformly over the admissible n; Chebyshev’s inequality then gives the required concentration. We will insert the explicit variance bound and the resulting (1+o(1)) preservation statement into the revised Section 2. revision: yes
Referee: [Theorem 1.2] Application to ℕ^k (Theorem 1.2 or equivalent): the claim that the weighted h-fold asymptotic holds for the full basis B = ℕ^k at h ≥ k² - k + O(√k) is load-bearing; a precise citation to the theorem establishing this asymptotic (including the precise form of the error term) must be supplied, as the subbasis existence reduces directly to it.

Authors: The threshold h ≥ k² - k + O(√k) is taken from the best available asymptotic formula for the weighted h-fold sum over all k-th powers. We will add the precise reference together with the exact statement of the main term and error term in the revised introduction and in the paragraph preceding Theorem 1.2. revision: yes

Circularity Check

0 steps flagged

No significant circularity; result conditional on external asymptotic for full basis

full rationale

The paper states a general probabilistic principle that is explicitly conditional on an assumed asymptotic for the weighted h-fold sum over the full basis B. It then applies the principle to B = ℕ^k (and ℙ^k) once h meets the stated threshold, yielding existence of subbases A with the prescribed r_{A,h}(n) ∼ 𝔖_{k,h}(n) F(n). No step in the provided text reduces the target growth F(n) or the existence claim back into the assumption by definition, fitting, or self-citation chain; the construction selects A probabilistically from the assumed main term on B without feeding the conclusion into the premise. The derivation remains self-contained against the stated external benchmark (the weighted asymptotic on B) and exhibits no self-definitional, fitted-input, or load-bearing self-citation patterns.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on one domain assumption: an asymptotic for the weighted h-fold representation sum over the full basis B. No free parameters or invented entities are visible in the abstract.

axioms (1)

domain assumption An asymptotic holds for a weighted h-fold representation sum over the basis B
Explicitly invoked as the premise that allows the probabilistic selection of the subbasis A with prescribed r_{A,h}.

pith-pipeline@v0.9.0 · 5775 in / 1418 out tokens · 33775 ms · 2026-05-23T05:00:16.352472+00:00 · methodology

discussion (0)

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Thin subbases of Piatetski-Shapiro sequences
math.NT 2026-05 unverdicted novelty 6.0

Piatetski-Shapiro sequences N_(c) contain thin subbases A of order h>=5 (for 1<c<2) or h>=(floor(2c)+1)(floor(2c)+2)+1 (for c>2), with r_{A,h}(n) ~ F(n) for regularly varying F satisfying the stated growth bounds.

Reference graph

Works this paper leans on

20 extracted references · 20 canonical work pages · cited by 1 Pith paper · 1 internal anchor

[1]

N. H. Bingham, C. M. Goldie, and J. L. Teugels, Regular variation, Cambridge Univ. Press, 1989

work page 1989
[2]

Br¨ udern and T

J. Br¨ udern and T. D. Wooley, On Waring’s problem for larger powers , J. Reine Angew. Math. 805 (2023), 115–142

work page 2023
[3]

Erd˝ os,Problems and results in additive number theory , Colloque sur la Theorie des Nombres (CBRM) (Bruxelles), 1956, pp

P . Erd˝ os,Problems and results in additive number theory , Colloque sur la Theorie des Nombres (CBRM) (Bruxelles), 1956, pp. 127–137

work page 1956
[4]

, Problems and results on additive properties of general sequ ences, II, Acta Math. Hung. 48 (1986), 127–137

work page 1986
[5]

Erd˝ os and P

P . Erd˝ os and P . Tetali,Representations of integers as the sum of k terms, Random Struct. Algor. 1 (1990), 245–261

work page 1990
[6]

Granville, Reﬁnements of Goldbach’s conjecture, and the generalized R iemann hypothesis, Funct

A. Granville, Reﬁnements of Goldbach’s conjecture, and the generalized R iemann hypothesis, Funct. Approx. Comment. Math. 37 (2007), 159–173

work page 2007
[7]

Harman, Trigonometric sums over primes, I , Mathematika 28 (1981), 249–254

G. Harman, Trigonometric sums over primes, I , Mathematika 28 (1981), 249–254

work page 1981
[8]

L. K. Hua, Additive theory of prime numbers , American Mathematical Society, 1965

work page 1965
[9]

J. H. Kim and V . H. Vu, Concentration of multivariate polynomials and its applica tions, Combinatorica 20 (2000), no. 3, 417–434

work page 2000
[10]

A. V . Kumchev and D. I. Tolev, An invitation to additive prime number theory , Serdica Math J. 31 (2005), 1–74

work page 2005
[11]

H. L. Montgomery and R. C. V aughan, Multiplicative number theory I: Classical theory , Cambridge Studies in Advanced Mathematics, vol. 97, Cambridge University Pre ss, Cambridge, 2006

work page 2006
[12]

M. B. Nathanson, Additive number theory: The classical bases , 2nd ed., Graduate Texts in Mathematics, vol. 164, Springer, 1996

work page 1996
[13]

Pliego, On Vu’s theorem in Waring’s problem for thinner sequences , arXiv:2410.11832, 2024

J. Pliego, On Vu’s theorem in Waring’s problem for thinner sequences , arXiv:2410.11832, 2024

work page arXiv 2024
[14]

Representation functions with prescribed rates of growth

C. T´ afula,Representation functions with prescribed rates of growth , arXiv:2405.01530, 2024

work page internal anchor Pith review Pith/arXiv arXiv 2024
[15]

R. C. V aughan, The Hardy–Littlewood method, 2nd ed., Cambridge Univ. Press, 1997

work page 1997
[16]

V . H. Vu, On a reﬁnement of Waring’s problem , Duke Math. J. 105 (2000), 107–134

work page 2000
[17]

, On the concentration of multivariate polynomials with smal l expectation, Random Struct. Algor. 16 (2000), 344–363

work page 2000
[18]

Wirsing, Thin subbases, Analysis 6 (1986), 285–308

E. Wirsing, Thin subbases, Analysis 6 (1986), 285–308

work page 1986
[19]

T. D. Wooley, On Vu’s thin basis theorem in Waring’s problem , Duke Math. J. 120 (2003), 1–34

work page 2003
[20]

, A light-weight version of Waring’s problem , J. Austral. Math. Soc. 76 (2004), 303–316. D ´EPARTEMENT DE MATH ´EMATIQUES ET DE STATISTIQUE , U NIVERSIT ´E DE MONTR ´EAL , CP 6128 SUCC . CENTRE -V ILLE , M ONTR ´EAL , QC H3C 3J7, C ANADA Email address: christian.tafula.santos@umontreal.ca

work page 2004

[1] [1]

N. H. Bingham, C. M. Goldie, and J. L. Teugels, Regular variation, Cambridge Univ. Press, 1989

work page 1989

[2] [2]

Br¨ udern and T

J. Br¨ udern and T. D. Wooley, On Waring’s problem for larger powers , J. Reine Angew. Math. 805 (2023), 115–142

work page 2023

[3] [3]

Erd˝ os,Problems and results in additive number theory , Colloque sur la Theorie des Nombres (CBRM) (Bruxelles), 1956, pp

P . Erd˝ os,Problems and results in additive number theory , Colloque sur la Theorie des Nombres (CBRM) (Bruxelles), 1956, pp. 127–137

work page 1956

[4] [4]

, Problems and results on additive properties of general sequ ences, II, Acta Math. Hung. 48 (1986), 127–137

work page 1986

[5] [5]

Erd˝ os and P

P . Erd˝ os and P . Tetali,Representations of integers as the sum of k terms, Random Struct. Algor. 1 (1990), 245–261

work page 1990

[6] [6]

Granville, Reﬁnements of Goldbach’s conjecture, and the generalized R iemann hypothesis, Funct

A. Granville, Reﬁnements of Goldbach’s conjecture, and the generalized R iemann hypothesis, Funct. Approx. Comment. Math. 37 (2007), 159–173

work page 2007

[7] [7]

Harman, Trigonometric sums over primes, I , Mathematika 28 (1981), 249–254

G. Harman, Trigonometric sums over primes, I , Mathematika 28 (1981), 249–254

work page 1981

[8] [8]

L. K. Hua, Additive theory of prime numbers , American Mathematical Society, 1965

work page 1965

[9] [9]

J. H. Kim and V . H. Vu, Concentration of multivariate polynomials and its applica tions, Combinatorica 20 (2000), no. 3, 417–434

work page 2000

[10] [10]

A. V . Kumchev and D. I. Tolev, An invitation to additive prime number theory , Serdica Math J. 31 (2005), 1–74

work page 2005

[11] [11]

H. L. Montgomery and R. C. V aughan, Multiplicative number theory I: Classical theory , Cambridge Studies in Advanced Mathematics, vol. 97, Cambridge University Pre ss, Cambridge, 2006

work page 2006

[12] [12]

M. B. Nathanson, Additive number theory: The classical bases , 2nd ed., Graduate Texts in Mathematics, vol. 164, Springer, 1996

work page 1996

[13] [13]

Pliego, On Vu’s theorem in Waring’s problem for thinner sequences , arXiv:2410.11832, 2024

J. Pliego, On Vu’s theorem in Waring’s problem for thinner sequences , arXiv:2410.11832, 2024

work page arXiv 2024

[14] [14]

Representation functions with prescribed rates of growth

C. T´ afula,Representation functions with prescribed rates of growth , arXiv:2405.01530, 2024

work page internal anchor Pith review Pith/arXiv arXiv 2024

[15] [15]

R. C. V aughan, The Hardy–Littlewood method, 2nd ed., Cambridge Univ. Press, 1997

work page 1997

[16] [16]

V . H. Vu, On a reﬁnement of Waring’s problem , Duke Math. J. 105 (2000), 107–134

work page 2000

[17] [17]

, On the concentration of multivariate polynomials with smal l expectation, Random Struct. Algor. 16 (2000), 344–363

work page 2000

[18] [18]

Wirsing, Thin subbases, Analysis 6 (1986), 285–308

E. Wirsing, Thin subbases, Analysis 6 (1986), 285–308

work page 1986

[19] [19]

T. D. Wooley, On Vu’s thin basis theorem in Waring’s problem , Duke Math. J. 120 (2003), 1–34

work page 2003

[20] [20]

, A light-weight version of Waring’s problem , J. Austral. Math. Soc. 76 (2004), 303–316. D ´EPARTEMENT DE MATH ´EMATIQUES ET DE STATISTIQUE , U NIVERSIT ´E DE MONTR ´EAL , CP 6128 SUCC . CENTRE -V ILLE , M ONTR ´EAL , QC H3C 3J7, C ANADA Email address: christian.tafula.santos@umontreal.ca

work page 2004