REVIEW 2 major objections 4 minor 32 references
Counterexamples of Friedlander--Iwaniec dual sums conjecture
T0 review · 2 major / 4 minor · reviewed 2026-08-01 · deepseek-v4-flash
Pith's one-line read The Friedlander–Iwaniec dual-sum conjecture fails for every zeta power of degree m≥4.
desk verdict A clean, elementary endpoint-resonance construction that likely refutes Friedlander–Iwaniec's uniform subpower bound for dual sums of ζ(s)^m, with the main uncertainty being the exact scope of the original conjecture. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The sharp dual sum B_{ℓ,D}(x,N) and a general endpoint-sector lemma (Theorem 2.1). For a phase Φ(n)=κ(nx)^ν with 0<ν<1, the derivative at n=N is of size N^{ν(a+1)-1} when x≍N^a, so the phase changes by only a bounded amount on a block of reciprocal-derivative length H≍N^{1-ν(a+1)}. Choosing x so the terminal phase hits a non-zero point of the periodic profile forces the whole block to align; with coefficients bounded below by c*>0, the block's weighted sum is power-sized. Applied to ζ(s)^m, the coefficients are d_m(n)≥1, ν=1/m, and the profile is the cosine.
What would settle it
Consult the original statement of the conjecture to verify that a degree-m Dirichlet series with a pole of order m≥4 at s=1 and κ_j=0 is explicitly admitted; alternatively, numerically compute B_4(x_N,N) for m=4, a=0.4, λ=1 for large N and check whether its magnitude consistently exceeds, say, N^{0.02} as predicted.
Extended reading notes
Core claim
Theorem 1.2 and Corollary 1.3: for every integer m≥4 there exist x_N≍N^a with 0<a<(m-3)/2 and M_N∈{N-H_N,N} such that |B_m(x_N,M_N)| ≫ N^{(m-3-2a)/(2m)}. Since the exponent is positive, the uniform subpower bound B_{ℓ,D}(x,N)≪(DNx)^ε in the Friedlander–Iwaniec conjecture fails for A(s)=B(s)=ζ(s)^m. The mechanism is endpoint resonance: choosing x so that the phase 2πm(nx)^{1/m}+φ_m is at a cosine maximum when n=N makes the phase vary by O(1) over a block of length H ≍ N^{1-(a+1)/m} preceding N, and all those terms contribute with the same sign.
Load-bearing premise
The counterexample applies only if ζ(s)^m, with its pole of order m at s=1 and boundary spectral parameters κ_j=0, is an admissible automorphic series in the original conjecture; the paper asserts this on the basis of the original framework, but if that framework secretly requires cuspidal L-functions with simple poles, the counterexample would lie outside its scope.
Editorial extensions
If this is right
- The uniform subpower estimate in Friedlander–Iwaniec Conjecture 1 is false for the zeta-power datum; any proof would have to exclude such poles or restrict the class of series.
- The counterexample satisfies N≤x for m≥6, so relaxing or enforcing that condition does not save the conjecture in the zeta-product case.
- The proposed deduction of the omega-free error-term conjecture (Conjecture 2) from Conjecture 1 via N≍x^{(m-1)/2} collapses for these examples, though Conjecture 2 itself is not disproved.
- Any sharp-cutoff sum with positive coefficients bounded below by a constant and a slowly varying nonlinear phase will exhibit the same resonance; the obstruction is not peculiar to zeta powers.
Reading between the lines
- The endpoint-sector lemma suggests that any L-function with a pole of order at least 4 at s=1 (or otherwise yielding coefficients bounded below) can be used to build counterexamples; one might test this for products of Dirichlet L-functions with multiple poles.
- The conjecture may remain plausible for cuspidal automorphic forms whose coefficients are not bounded below by a positive constant (e.g., Hecke eigenvalues with sign changes); the lower bound d_m(n)≥1 is essential to the construction.
- A quick numerical check for m=4, a=0.4, N=10^6 using the explicit formula (3.4) should show |B_4| growing like N^{1/40}, which would settle the matter concretely.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper claims to disprove the Friedlander–Iwaniec sharp dual-sum Conjecture 1 by constructing explicit power-sized values of the sharply truncated dual sum for A(s)=B(s)=ζ(s)^m, m≥4. The core is an elementary endpoint-sector lemma (Theorem 2.1): for any phase κ(nx)^ν with 0<ν<1 and coefficients bounded below, one can choose x_N ≍ N^a so that the terminal H_N ≍ N^{1−ν(a+1)} terms lie in a fixed arc where the periodic weight has positive real part, forcing a block contribution ≫ N^{1−β−ν(a+1)}. Specialized with ν=1/m, β=(m+1)/(2m), κ=2πm and c_n=d_m(n), this yields Theorem 1.2 and Corollary 1.3: for m≥4 and 0<a<(m−3)/2, the dual sum is ≫ N^{(m−3−2a)/(2m)} along a sequence, contradicting the conjectured uniform subpower bound. The paper also gives an N≤x counterexample for m≥6 and an N ≍ x^{(m−1)/2} counterexample for m≥5.
Significance. If the Friedlander–Iwaniec conjecture is correctly interpreted to include ζ(s)^m with pole order m≥4 and arbitrary N, the paper's result is a clean, explicit refutation of a published conjecture. The proof is transparent and elementary, with no fitted constants, machine-verifiable computations, and explicit sequences; it also explains why the trivial bound in (1.10) is the best possible in this case. The paper's self-imposed limitations (Remark 2.3, Section 1.5) show awareness of the boundaries of the argument. The main risk is external: the scope of the original conjecture is not quoted, and the counterexample relies on a specific reading of [6]. The m≥6, N≤x part of Corollary 1.3(ii) is robust to the N≤x restriction.
major comments (2)
- [Section 1.3, Conjecture 1.1] The conjecture is presented as a paraphrase without quoting the original. The paper's claims that [6, pp. 494–495] explicitly allow products of zeta with poles of arbitrary finite order, and that (1.9) is uniform in N without the restriction 1≤N≤x, are load-bearing. If the original conjecture restricts N≤x, the m=4,5 counterexamples (with a<1 so N ≫ x_N) are outside the conjecture's range, leaving only m≥6 with N≤x. If the class of "automorphic series" is narrower than the paper's reading (e.g., cuspidal factors with simple poles), ζ(s)^m may not be a permissible datum. Please quote the exact statement of [6, Conjecture 1] and the cited passage, and amend the abstract/theorems if any restriction applies.
- [Section 3.1, Proposition 3.1] The verification that ζ(s)^m lies in the Friedlander–Iwaniec framework relies on the assertion "Re κ_j=0 is explicitly permitted in [6, (1.4)]" and "pole at s=1 of arbitrary finite order [6, pp. 494–495]." These are not demonstrated in the manuscript. Because the counterexample's validity depends entirely on this inclusion, the authors should reproduce the relevant conditions verbatim and confirm that the boundary case κ_j=0 and pole order m are allowed. If they are not, the counterexample does not lie in the class covered by the conjecture.
minor comments (4)
- [Section 1.6] Typo: "Organiztion" should be "Organization."
- [Abstract] Grammar: "Let a(n) and b(n) are arithmetic sequences" should be "Let a(n) and b(n) be arithmetic sequences."
- [Section 1.3] The condition "1≤N≤x" appears in (1.8) but is omitted in Conjecture 1.1; clarify whether the conjecture is stated for all N or for the same range, and ensure the notation B_{\ell,D}(x,N) is consistent.
- [Remark 2.3] The statement "This is only an amplitude-level statement..." is useful, but it might be placed earlier to avoid confusion about the role of Clozel–Sarnak's framework.
Circularity Check
No circularity: the endpoint construction is self-contained and the conjecture is the target, not an input.
full rationale
The derivation chain is self-contained. Theorem 2.1 is an elementary endpoint-sector lemma proved directly from a nonvanishing periodic profile, a positive coefficient lower bound, and a first-derivative scale; it does not presuppose the conjecture. Theorem 1.2 applies it to ζ(s)^m by way of Proposition 3.1, which verifies the Friedlander–Iwaniec normalization using only the classical zeta functional equation, the coefficient bound d_m(n) ≪ n^ε, and the quoted conditions (3.1)–(3.2). No fitted parameter is renamed as a prediction: the resonant x_N is explicitly constructed so that the terminal phase is exactly a cosine maximum, and the lower bound (1.16) measures the coherent terminal block. Conjecture 1.1 is the target of the refutation; the contradiction argument assumes it momentarily and derives its failure, which is a standard refutation rather than a circular use of the conjecture as an input. There are no self-citations by this author. The only load-bearing external dependency is the scope of the Friedlander–Iwaniec conjecture as actually stated in [6] — whether ζ(s)^m with pole order m ≥ 4 is an admissible automorphic series and whether the bound is required for N > x. That is a factual or interpretational correctness risk, and the paper itself flags related limitations in Remark 2.3 and Section 1.5, but it is not circular reasoning. Accordingly the circularity score is 0.
Assumptions & free parameters
assumptions (3)
- standard math ζ(s) has meromorphic continuation to C with a unique simple pole at s=1 and satisfies ζ(1-s)=π^{1/2-s}Γ(s/2)/Γ((1-s)/2)ζ(s); hence ζ(s)^m has a pole of order m and degree-m functional equation.
- domain assumption The Friedlander–Iwaniec framework admits products of zeta and Dirichlet L-functions and allows poles of arbitrary finite order at s=1, with boundary spectral parameters Re κ_j=0 permitted.
- domain assumption For a real archimedean parameter array, the two exponential height modes in the Friedlander–Iwaniec summation formula combine into the single cosine in (1.7) with ℓ=m-3.
Cite this review
Pith. "Pith review of Counterexamples of Friedlander--Iwaniec dual sums conjecture." pith.science (2026). https://pith.science/paper/PCF2GGCY
@misc{pith2026260716695,
author = {Pith},
title = {Pith review of: Counterexamples of Friedlander--Iwaniec dual sums conjecture},
year = {2026},
howpublished = {\url{https://pith.science/paper/PCF2GGCY}},
note = {Machine review of arXiv:2607.16695}
}
abstract
Let $a(n)$ and $b(n)$ be arithmetic sequences, and $$A(s)=\sum_{n\ge1}a(n)n^{-s}, \qquad B(s)=\sum_{n\ge1}b(n)n^{-s},$$ be the two Dirichlet series related by a certain functional equation. Let $m$ be the \emph{analytic degree} of the functional equation. For $x>0$ and a positive integer $N$, Friedlander and Iwaniec (2005) define the sharply truncated nonlinear dual sum $$\mathcal B_{\ell,D}(x,N) := \sum_{\substack{n\in\mathbb N\\ n\le N}} b(n)n^{-\beta_m} \cos\left( 2\pi m\left(\frac{nx}{D}\right)^{1/m} +\frac{\pi\ell}{4} \right),$$ where $D\ge1$ is the conductor, $\beta_m:=\frac{m+1}{2m}$, and $\ell=m-3-2k$ is determined by the archimedean weight $k$ of the functional equation. Their Conjecture 1 predicts that, for every $\varepsilon>0$, $$\mathcal B_{\ell,D}(x,N) \ll_{\varepsilon,\boldsymbol\kappa} (DNx)^\varepsilon,$$ uniformly in the variables $x$ and $N$, with the degree, conductor, and archimedean datum fixed. We give counterexamples to this prediction with $$A(s)=B(s)=\zeta(s)^m,\; m\geq 4$$ where $$\zeta(s):=\sum_{n\ge1}n^{-s} \qquad(\operatorname{Re}s>1)$$ is the Riemann zeta function.
Reference graph
Works this paper leans on
-
[6]
J. B. Friedlander and H. Iwaniec,Summation formulae for coefficients ofL-functions, Canad. J. Math.57(2005), no. 3, 494–505. doi:10.4153/CJM-2005-021-5
-
[1]
J. Arias de Reyna,On the Kuzmin–Landau lemma, arXiv:2002.05982, 2020. doi: 10.48550/arXiv.2002.05982
work page Pith review arXiv doi:10.48550/arxiv.2002.05982 2002
-
[2]
K. Chandrasekharan and R. Narasimhan,Functional equations with multiple gamma factors and the average order of arithmetical functions, Ann. of Math. (2)76(1962), 93–136. doi:10.2307/1970267
-
[3]
L. Clozel and P. Sarnak,A universal lower bound for certain quadratic integrals of automorphic L-functions, J. Number Theory261(2024), 252–298. doi:10.1016/j.jnt. 2024.02.018
doi:10.1016/j.jnt 2024
-
[4]
NISTDigital Libraryof MathematicalFunctions,Asymptotic Approximations, §2.3(iv)– (v), F. W. J. Olver et al., eds., https://dlmf.nist.gov/2.3
-
[5]
A.-M. Ernvall-Hytönen, J. Jääsaari, and E. V. Vesalainen,Resonances andΩ-results for exponential sums related to Maass forms forSL(n,Z ), J. Number Theory153 (2015), 135–157. doi:10.1016/j.jnt.2015.01.014
-
[7]
Jääsaari,Ω-results for exponential sums related to Maass cusp forms for SL3(Z), Math
J. Jääsaari,Ω-results for exponential sums related to Maass cusp forms for SL3(Z), Math. Proc. Cambridge Philos. Soc.180(2026), 531–584. doi:10.1017/ S0305004125101813
2026
-
[8]
J. L. Hafner and A. Ivić,On sums of Fourier coefficients of cusp forms, Enseign. Math. (2)35(1989), no. 3–4, 375–382. doi:10.5169/seals-57381
Show all 32 references
-
[9]
G. H. Hardy,On Dirichlet’s divisor problem, Proc. London Math. Soc. (2)15(1916), 1–25. doi:10.1112/plms/s2-15.1.1
1916 doi
-
[10]
M. N. Huxley,Exponential sums and lattice points. III, Proc. London Math. Soc. (3) 87(2003), no. 3, 591–609. doi:10.1112/S0024611503014485
2003 doi
-
[11]
Iwaniec and E
H. Iwaniec and E. Kowalski,Analytic Number Theory, Amer. Math. Soc. Colloq. Publ. 53, American Mathematical Society, Providence, RI, 2004
2004
-
[12]
Kaczorowski and A
J. Kaczorowski and A. Perelli,The standard twist ofL-functions revisited, Acta Arith. 201(2021), no. 3, 281–328. doi:10.4064/aa210730-23-8
2021 doi
-
[13]
Kolesnik,On the estimation of multiple exponential sums, in: Recent Progress in Analytic Number Theory, Vol
G. Kolesnik,On the estimation of multiple exponential sums, in: Recent Progress in Analytic Number Theory, Vol. I (Durham, 1979), Academic Press, London–New York, 1981, 231–246
1979
-
[14]
Lamzouri,Sharp omega results for the divisor and circle problems, arXiv:2605.21476,
Y. Lamzouri,Sharp omega results for the divisor and circle problems, arXiv:2605.21476,
-
[15]
Leung,A central limit theorem for coefficients ofL-functions in short intervals, Bull
S.-K. Leung,A central limit theorem for coefficients ofL-functions in short intervals, Bull. London Math. Soc.57(2025), no. 3, 831–853. doi:10.1112/blms.70002
2025 doi
- [16]
-
[17]
Lin and Q
Y. Lin and Q. Sun,Analytic twists ofGL3×GL 2 automorphic forms, Int. Math. Res. Not. IMRN2021(2021), no. 19, 15143–15208. doi:10.1093/imrn/rnaa348
2021 doi
-
[18]
Mahatab,Omega results for the divisor and circle problems using the resonance method, Ramanujan J.69(2026), no
K. Mahatab,Omega results for the divisor and circle problems using the resonance method, Ramanujan J.69(2026), no. 4, article 91. doi:10.1007/s11139-026-01368-8
2026 doi
-
[19]
Matomäki and X
K. Matomäki and X. Shao,Discorrelation between primes in short intervals and polynomial phases, Int. Math. Res. Not. IMRN2021(2021), no. 16, 12330–12355. doi:10.1093/imrn/rnz188
2021 doi
-
[20]
S. D. Miller and W. Schmid,Summation formulas, from Poisson and Voronoi to the present, in: Noncommutative Harmonic Analysis, Progr. Math.220, Birkhäuser Boston, Boston, MA, 2004, 419–440. doi:10.1007/978-0-8176-8204-0_15
2004 doi
-
[21]
S. D. Miller and W. Schmid,Automorphic distributions,L-functions, and Voronoi summation for GL(3), Ann. of Math. (2)164(2006), no. 2, 423–488. doi:10.4007/ annals.2006.164.423
2006
-
[22]
S. D. Miller and W. Schmid,A general Voronoi summation formula forGL(n,Z ), in: Geometry and Analysis. No. 2, Adv. Lect. Math. (ALM)18, International Press, Somerville, MA, 2011, 173–224
2011
-
[23]
Ren and Y.-B
X.-M. Ren and Y.-B. Ye,Resonance between automorphic forms and exponential func- tions, Sci. China Math.53(2010), no. 9, 2463–2472. doi:10.1007/s11425-010-3150-4
2010 doi
-
[24]
Ren and Y.-B
X.-M. Ren and Y.-B. Ye,Resonance and rapid decay of exponential sums of Fourier coefficients of a Maass form forGLm(Z), Sci. China Math.58(2015), no. 10, 2105–
2015
-
[25]
Soundararajan,Ωresults for the divisor and circle problems, Int
K. Soundararajan,Ωresults for the divisor and circle problems, Int. Math. Res. Not. 2003(2003), no. 36, 1987–1998. doi:10.1155/S1073792803130309
2003 doi
-
[26]
Sourmelidis,An additive application of the resonance method, Mathematika72 (2026), no
A. Sourmelidis,An additive application of the resonance method, Mathematika72 (2026), no. 3, e70106. doi:10.1112/mtk.70106
2026 doi
-
[27]
E. C. Titchmarsh,The Theory of the Riemann Zeta-Function, 2nd ed., revised by D. R. Heath-Brown, Clarendon Press, Oxford, 1986
1986
-
[28]
Voronoï,Sur une fonction transcendante et ses applications à la sommation de quelques séries, Ann
G. Voronoï,Sur une fonction transcendante et ses applications à la sommation de quelques séries, Ann. Sci. Ècole Norm. Sup. (3)21(1904), 207–267. doi:10.24033/ asens.539
1904
-
[29]
Voronoï,Sur une fonction transcendante et ses applications à la sommation de quelques séries (suite), Ann
G. Voronoï,Sur une fonction transcendante et ses applications à la sommation de quelques séries (suite), Ann. Sci. Ècole Norm. Sup. (3)21(1904), 459–533. doi: 10.24033/asens.545
1904 doi
-
[30]
Wu,Power sums of Hecke eigenvalues and application, Acta Arith.137(2009), no
J. Wu,Power sums of Hecke eigenvalues and application, Acta Arith.137(2009), no. 4, 333–344. doi:10.4064/aa137-4-3. Morningside Center of Mathematics, Chinese Academy of Sciences, No. 55 Zhongguancun East Road, Beijing 100190, China Email address:khaihoann@gmail.com
2009 doi
- [2026]
-
[2124]
doi:10.1007/s11425-014-4955-3
Reviewed August 1, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.