The asymptotic in Waring's problem over function fields via a singular locus in the circle method

Will Sawin

arxiv: 2412.14053 · v3 · submitted 2024-12-18 · 🧮 math.NT

The asymptotic in Waring's problem over function fields via a singular locus in the circle method

Will Sawin This is my paper

Pith reviewed 2026-05-23 06:44 UTC · model grok-4.3

classification 🧮 math.NT

keywords Waring's problemfunction fieldscircle methodsingular locusexponential sumsKatz boundsManin's conjectureFermat hypersurfaces

0 comments

The pith

Treating minor arcs as finite-field exponential sums and bounding them via the dimension of a singular locus yields stronger asymptotics for Waring's problem over function fields than the integer case.

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

The paper shows that Waring's problem over function fields admits asymptotic formulas with error terms better than those available over the integers from the main conjecture on Vinogradov's mean value theorem. Instead of analytic estimates on the minor arcs, the argument converts those arcs into complete exponential sums over finite fields and invokes Katz's theorems that bound such sums in terms of the dimension of an associated singular locus. Tangent-space calculations supply the needed dimension bound. The same technique produces corresponding improvements for the asymptotic count of rational points on Fermat hypersurfaces in the setting of Manin's conjecture.

Core claim

Estimating the dimension of the singular locus that appears in the exponential sum for the circle method over function fields, via direct tangent-space calculations, produces bounds on the sum that are strong enough to give an asymptotic for the number of representations in Waring's problem stronger than the best known results over the integers; the identical method yields improved estimates for Manin's conjecture on Fermat hypersurfaces over function fields.

What carries the argument

Katz's bound on complete exponential sums over finite fields, controlled by the dimension of a singular locus whose dimension is computed by tangent-space calculations.

If this is right

The asymptotic formula for representations in Waring's problem over function fields has a smaller error term than the corresponding formula over the integers.
The same singular-locus method supplies stronger error terms for the asymptotic count of rational points on Fermat hypersurfaces, relevant to Manin's conjecture.
Analytic estimates on minor arcs can be replaced by algebraic bounds coming from Katz's results once the singular locus dimension is controlled.
Tangent-space calculations are sufficient to determine the dimension of the relevant singular locus in these problems.

Where Pith is reading between the lines

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

The method may extend to other Diophantine problems over function fields where exponential sums arise and a singular locus can be described geometrically.
The gain over the integer case illustrates how algebraic-geometry tools available only in positive characteristic can sometimes outpace the best analytic tools in characteristic zero.
If the tangent-space technique works here, similar dimension estimates could be attempted for exponential sums attached to other hypersurface equations.

Load-bearing premise

Tangent-space calculations give a sufficiently tight upper bound on the dimension of the singular locus to produce the stated improvement over the integer bounds.

What would settle it

An explicit example or computation in which the actual dimension of the singular locus exceeds the tangent-space upper bound used in the argument would remove the claimed improvement.

read the original abstract

We give results on the asymptotic in Waring's problem over function fields that are stronger than the results obtained over the integers using the main conjecture in Vinogradov's mean value theorem. Similar estimates apply to Manin's conjecture for Fermat hypersurfaces over function fields. Following an idea of Pugin, rather than applying analytic methods to estimate the minor arcs, we treat them as complete exponential sums over finite fields and apply results of Katz, which bound the sum in terms of the dimension of a certain singular locus, which we estimate by tangent space calculations.

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

Sawin gets stronger Waring asymptotics over function fields than the Vinogradov main conjecture delivers over integers by replacing minor-arc analysis with Katz bounds on complete sums, where the gain comes from tangent-space estimates of singular-locus dimension.

read the letter

The core advance is that the paper produces an asymptotic for the number of representations in Waring's problem over F_q(t) that improves on what the main conjecture in Vinogradov's mean value theorem would yield in the integer setting, and it carries the same method over to Manin's conjecture for Fermat hypersurfaces. The method follows Pugin's suggestion: treat the minor arcs as complete exponential sums over finite fields, invoke Katz's bound in terms of the dimension of a singular locus, and compute that dimension via tangent-space calculations at candidate points. This is a genuine transfer of algebraic-geometry input into the circle method that is not present in the integer literature cited in the abstract. The execution looks direct and avoids extra parameters or fitted constants. The obvious soft spot is whether the tangent-space upper bounds on dim Sing are tight enough to clear the threshold needed for the claimed improvement. Katz's estimate is roughly q to the power of (n + dim Sing)/2, so any inflation in the dimension bound immediately weakens the final exponent. The abstract indicates the calculations are done pointwise, but without seeing the rank computations or checks for reducedness and equidimensionality it is impossible to confirm the bound is strict enough. Minor issues such as the precise range of degrees or the size of the implied constant are secondary. The paper is aimed at people who already work with the circle method in positive characteristic or with function-field analogs of Manin-type problems. A reader who cares about explicit error terms in these settings will find concrete new estimates to compare against the integer case. It deserves a serious referee because the claims are specific, the method is reproducible in principle, and the potential gain is stated clearly enough to be checked or refuted by direct calculation.

Referee Report

2 major / 2 minor

Summary. The paper proves asymptotic formulas for Waring's problem over the function field F_q(t) that improve on the error terms obtainable over Z even under the main conjecture in Vinogradov's mean value theorem. Minor arcs are handled as complete exponential sums over finite fields; Katz's theorems bound these sums in terms of the dimension of a singular locus on an auxiliary variety, which the authors estimate by explicit tangent-space calculations. The same method yields corresponding statements for Manin's conjecture on Fermat hypersurfaces over function fields.

Significance. If the dimension bounds are verified to be sufficiently sharp, the work supplies unconditional results in the function-field setting that surpass the best conjectural integer bounds and demonstrates the utility of algebraic-geometry tools (Katz's exponential-sum estimates) for Diophantine problems over global fields of positive characteristic. The approach avoids reliance on the still-open VMVT conjecture and produces falsifiable predictions for the function-field case.

major comments (2)

[§3] §3 (singular-locus dimension estimate): the tangent-space rank calculations are asserted to produce dim Sing(V) small enough to beat the threshold that would recover only the conjectural integer exponent. The manuscript must explicitly state the numerical threshold required for the claimed improvement, verify that the bound is strict at every point of the candidate locus, and confirm that the locus is reduced and equidimensional so that Katz's theorem applies with the stated exponent.
[§4] §4 (application of Katz): the precise form of the bound |sum| ≪ q^{(n + dim Sing)/2 + O(1)} is invoked, but the error term O(1) and the precise dependence on the degree must be tracked through to the final asymptotic to ensure the claimed saving is realized.

minor comments (2)

[Introduction and §3] Notation for the auxiliary variety and its singular locus should be introduced once and used consistently; currently the same symbol appears with two different meanings in the introduction and in §3.
[Introduction] The reference to Pugin's idea is cited only in the abstract; a precise pointer to the relevant passage in Pugin's work should be added in the introduction.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for their detailed and insightful report. Their comments have helped us clarify several aspects of the singular locus estimates and the application of Katz's theorems. We address each major comment below and have revised the manuscript accordingly to incorporate the suggested improvements.

read point-by-point responses

Referee: [§3] §3 (singular-locus dimension estimate): the tangent-space rank calculations are asserted to produce dim Sing(V) small enough to beat the threshold that would recover only the conjectural integer exponent. The manuscript must explicitly state the numerical threshold required for the claimed improvement, verify that the bound is strict at every point of the candidate locus, and confirm that the locus is reduced and equidimensional so that Katz's theorem applies with the stated exponent.

Authors: We thank the referee for highlighting the need for explicit verification in §3. In the revised version of the manuscript, we have explicitly stated the numerical threshold required for the claimed improvement over the conjectural bounds from Vinogradov's mean value theorem. The tangent-space rank calculations have been presented in a manner that verifies the dimension bound is strict at every point of the candidate locus. We have also confirmed that the singular locus is reduced and equidimensional, so that Katz's theorem applies with the stated exponent. revision: yes
Referee: [§4] §4 (application of Katz): the precise form of the bound |sum| ≪ q^{(n + dim Sing)/2 + O(1)} is invoked, but the error term O(1) and the precise dependence on the degree must be tracked through to the final asymptotic to ensure the claimed saving is realized.

Authors: Regarding the application of Katz's bound in §4, we have now explicitly tracked the O(1) error term, which is independent of q, and the dependence on the degree of the hypersurface through the estimates. This tracking confirms that the claimed saving in the asymptotic formula is realized without being affected by these terms. The revised manuscript includes the detailed tracking in the minor arcs analysis. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation uses external Katz bounds and independent tangent-space computations

full rationale

The paper obtains the asymptotic for Waring's problem over function fields by treating minor arcs as complete exponential sums and invoking Katz's theorem to bound them in terms of the dimension of a singular locus on an auxiliary variety; this dimension is then computed directly via tangent-space rank calculations at candidate points. These geometric computations are algebraic and do not presuppose or reduce to the target asymptotic formula. No parameters are fitted to the Waring count, no self-citation chain supplies a load-bearing uniqueness or ansatz, and the cited results of Katz and Pugin are external. The derivation chain therefore remains self-contained against independent algebraic-geometry input.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper rests on Katz's theorems bounding exponential sums by singular-locus dimension and on the standard algebraic-geometry fact that tangent spaces compute local dimensions; no free parameters or new invented entities are introduced in the abstract.

axioms (2)

standard math Katz's bounds on complete exponential sums in terms of the dimension of the associated singular locus hold over finite fields.
Invoked in the abstract as the replacement for analytic minor-arc estimates.
domain assumption Tangent-space calculations correctly determine the dimension of the singular locus arising from the Waring equation.
The abstract states that this estimation step produces the claimed bounds.

pith-pipeline@v0.9.0 · 5610 in / 1477 out tokens · 24569 ms · 2026-05-23T06:44:15.194569+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

12 extracted references · 12 canonical work pages

[1]

Bourgain

J. Bourgain. On the Vinogradov mean value. Proceedings of the Steklov Institute of Mathematics , 296(1):30–40, January 2017. doi:10.1134/S0081543817010035

work page doi:10.1134/s0081543817010035 2017
[2]

Proof of the m ain conjecture in Vinogradov’s mean value theorem for degrees higher than three

Jean Bourgain, Ciprian Demeter, and Larry Guth. Proof of the m ain conjecture in Vinogradov’s mean value theorem for degrees higher than three. Annals of Mathematics , 184(2):633–682, September 2016. doi:10.4007/annals.2016.184.2.7

work page doi:10.4007/annals.2016.184.2.7 2016
[3]

A higher genus circle method and an application to geometric Manin’s conjecture

Matthew Hase-Liu. A higher genus circle method and an application to geometric Manin’s conjecture. arXiv:2402.10498, 2024

work page arXiv 2024
[4]

singular

Nicholas Katz. Estimates for “singular” exponential sums. International Mathematics Research Notices , 1999(16):875, 1999. doi:10.1155/S1073792899000458

work page doi:10.1155/s1073792899000458 1999
[5]

Nicholas M. Katz. Sums of Betti numbers in arbitrary character istic. Finite Fields and Their Applica- tions, 7(1):29–44, January 2001. doi:10.1006/ﬀta.2000.0303. W ARING’S PROBLEM OVER FUNCTION FIELDS 45

work page arXiv 2001
[6]

R. M. Kubota. Waring’s problem for Fq[x]. Instytut Matematyczny Polskiej Akademi Nauk, 1974

work page 1974
[7]

Yu-Ru Liu and Trevor D. Wooley. Waring’s problem in function ﬁelds. Journal f¨ ur die reine und ange- wandte Mathematik (Crelles Journal) , 2010(638):1–67, January 2010. doi:10.1515/CRELLE.2010.001

work page doi:10.1515/crelle.2010.001 2010
[8]

An Algebraic Circle Method

Pugin, Thibaut. An Algebraic Circle Method . PhD thesis, Columbia University, 2011. doi:10.7916/D8G166VP

work page doi:10.7916/d8g166vp 2011
[9]

R.C. Vaughan. On Waring’s problem for cubes. Journal f¨ ur die reine und angewandte Mathematik , 365:122–170, 1986. http://eudml.org/doc/152808

work page 1986
[10]

Trevor D. Wooley. The asymptotic formula in Waring’s problem. International Mathematics Research Notices, 2012(7):1485–1504, 2012. doi:10.1093/imrn/rnr074

work page doi:10.1093/imrn/rnr074 2012
[11]

Trevor D. Wooley. Nested eﬃcient congruencing and relatives o f Vinogradov’s mean value theorem. Pro- ceedings of the London Mathematical Society , 118(4):942–1016, October 2018. doi:10.1112/plms.12204

work page doi:10.1112/plms.12204 2018
[12]

The asymptotic formula for Waring’s proble m in function ﬁelds

Shuntaro Yamagishi. The asymptotic formula for Waring’s proble m in function ﬁelds. International Mathematics Research Notices , 2016(23):7137–7178, March 2016. doi:10.1093/imrn/rnv392. Appendix A. Linear programming for the arbitrary hypersurface case Corollary A.1. The bound (25) is satisﬁed for δ >0 as long as n >5d − 5 + (d − 1)δ, e is suﬃciently larg...

work page doi:10.1093/imrn/rnv392 2016

[1] [1]

Bourgain

J. Bourgain. On the Vinogradov mean value. Proceedings of the Steklov Institute of Mathematics , 296(1):30–40, January 2017. doi:10.1134/S0081543817010035

work page doi:10.1134/s0081543817010035 2017

[2] [2]

Proof of the m ain conjecture in Vinogradov’s mean value theorem for degrees higher than three

Jean Bourgain, Ciprian Demeter, and Larry Guth. Proof of the m ain conjecture in Vinogradov’s mean value theorem for degrees higher than three. Annals of Mathematics , 184(2):633–682, September 2016. doi:10.4007/annals.2016.184.2.7

work page doi:10.4007/annals.2016.184.2.7 2016

[3] [3]

A higher genus circle method and an application to geometric Manin’s conjecture

Matthew Hase-Liu. A higher genus circle method and an application to geometric Manin’s conjecture. arXiv:2402.10498, 2024

work page arXiv 2024

[4] [4]

singular

Nicholas Katz. Estimates for “singular” exponential sums. International Mathematics Research Notices , 1999(16):875, 1999. doi:10.1155/S1073792899000458

work page doi:10.1155/s1073792899000458 1999

[5] [5]

Nicholas M. Katz. Sums of Betti numbers in arbitrary character istic. Finite Fields and Their Applica- tions, 7(1):29–44, January 2001. doi:10.1006/ﬀta.2000.0303. W ARING’S PROBLEM OVER FUNCTION FIELDS 45

work page arXiv 2001

[6] [6]

R. M. Kubota. Waring’s problem for Fq[x]. Instytut Matematyczny Polskiej Akademi Nauk, 1974

work page 1974

[7] [7]

Yu-Ru Liu and Trevor D. Wooley. Waring’s problem in function ﬁelds. Journal f¨ ur die reine und ange- wandte Mathematik (Crelles Journal) , 2010(638):1–67, January 2010. doi:10.1515/CRELLE.2010.001

work page doi:10.1515/crelle.2010.001 2010

[8] [8]

An Algebraic Circle Method

Pugin, Thibaut. An Algebraic Circle Method . PhD thesis, Columbia University, 2011. doi:10.7916/D8G166VP

work page doi:10.7916/d8g166vp 2011

[9] [9]

R.C. Vaughan. On Waring’s problem for cubes. Journal f¨ ur die reine und angewandte Mathematik , 365:122–170, 1986. http://eudml.org/doc/152808

work page 1986

[10] [10]

Trevor D. Wooley. The asymptotic formula in Waring’s problem. International Mathematics Research Notices, 2012(7):1485–1504, 2012. doi:10.1093/imrn/rnr074

work page doi:10.1093/imrn/rnr074 2012

[11] [11]

Trevor D. Wooley. Nested eﬃcient congruencing and relatives o f Vinogradov’s mean value theorem. Pro- ceedings of the London Mathematical Society , 118(4):942–1016, October 2018. doi:10.1112/plms.12204

work page doi:10.1112/plms.12204 2018

[12] [12]

The asymptotic formula for Waring’s proble m in function ﬁelds

Shuntaro Yamagishi. The asymptotic formula for Waring’s proble m in function ﬁelds. International Mathematics Research Notices , 2016(23):7137–7178, March 2016. doi:10.1093/imrn/rnv392. Appendix A. Linear programming for the arbitrary hypersurface case Corollary A.1. The bound (25) is satisﬁed for δ >0 as long as n >5d − 5 + (d − 1)δ, e is suﬃciently larg...

work page doi:10.1093/imrn/rnv392 2016