An analogue of Hirose's relation for finite multiple harmonic $q$-series at roots of unity

Takumi Maesaka

An analogue of Hirose's relation for finite multiple harmonic $q$-series at roots of unity

T0 review · 0 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read This paper proves a finite, root-of-unity analogue of a recent relation for refined symmetric multiple zeta values and uses it to prove the cyclic sum conjecture.

desk verdict Proves the q-analogue of Hirose's relation and settles the HPHPT cyclic sum conjecture; the proof is sound, with a terse continuity step that can be filled. read the letter →

arxiv 2608.09806 v1 pith:RJFWISTK submitted 2026-08-10 math.NT

Takumi Maesaka This is my paper · ORCID

classification math.NT MSC 11M3211R1805A30

keywords finitemultipleharmonicq-seriesrootsofunitycyclicsumformulaconnectedsumsq-harmonicproductzetavaluesrefinedsymmetric

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper establishes a finite, root-of-unity analogue of a recently proved relation among refined symmetric multiple zeta values. The main theorem asserts that for non-empty indices $k$ and $l$ with first and last entries at least $2$, the finite sum $z_N((\downarrow(k*_{\zeta_N} l)\downarrow)^\vee)$ is zero, where $*_{\zeta_N}$ is the $q$-harmonic product evaluated at a primitive $N$-th root of unity and the arrows and $\vee$ are standard index operations. Because the same $z_N$ family interpolates between refined symmetric multiple zeta values and finite multiple zeta values, the identity recovers the known symmetric relation as $N\to\infty$ and yields a finite version after reducing modulo the prime ideal generated by $1-\zeta_p$. The paper then proves a cyclic sum conjecture for finite multiple harmonic $q$-series at roots of unity, obtaining explicit evaluations that specialize to symmetric and finite cyclic sum formulas.

What carries the argument

The central object is the two-variable connected sum $Z_{q,s,t}(k;l)$ with connector $C(n,m)$; its transport relations $Z_{q,s,t}(k\to;l)=Z_{q,s,t}(k;l\uparrow)$ and the mirror identity give the duality formula $\zeta_q(k;s,t)=\zeta_q(k^\dagger;t,s)$. The paper extends this duality by continuity to boundary values, pairs it with the $q$-harmonic product $*_{q}$, and rewrites the vanishing identity using the Hoffman dual and up/down arrows. The argument closes by comparing coefficients in the topological basis $\{(q;q)_r t^r/(tq;q)_r\}$ of $\mathbb{C}[[q]][[t]]$ and letting $q\to\zeta_N$.

What would settle it

Compute $z_N((\downarrow(k*_{\zeta_N} l)\downarrow)^\vee)$ directly from the defining sum for a small admissible pair, for example $k=(2,2)$ and $l=(2)$ at $N=5$; the theorem predicts $0$, so a single nonzero value would disprove the central claim.

Watch

Extended reading notes

Core claim

The central claim is that the connected-sum duality behind the refined-symmetric relation survives evaluation at a root of unity. Theorem 1.2 states $z_N((\downarrow(k*_{\zeta_N} l)\downarrow)^\vee)=0$ for all non-empty indices $k$ and $l$ whose first and last entries are at least $2$. The proof uses two-variable connected sums $Z_{q,s,t}(k;l)$, their transport relations, and the duality formula $\zeta_q(k;s,t)=\zeta_q(k^\dagger;t,s)$, extended by continuity to the boundary $s=1$ or $t=1$. Expanding in a $q$-series basis and taking $q\to\zeta_N$ forces the relevant coefficient $z_N$ to vanish. With a quasi-shuffle identity and the single-index values $z_N(\{1\}^{j-2})$, the same result implies the cyclic sum formula and its symmetric and finite specializations.

Load-bearing premise

The proof relies on extending the duality identity for the $q$-series from the open domain $|q|,|s|,|t|<1$ to the boundary points $s=1$ or $t=1$ by continuity, and then evaluating at $q=\zeta_N$; if that boundary extension fails to converge absolutely for the admissible indices used, the coefficient comparison that produces $z_N$ would break.

Editorial extensions

If this is right

Letting $N\to\infty$ in Theorem 1.2 recovers the known symmetric multiple zeta value relation, giving an independent proof of it.
Setting $N=p$ and reducing modulo the ideal $(1-\zeta_p)$ yields the finite multiple zeta value version $\zeta_A((\downarrow(k*l)\downarrow)^\vee)=0$.
Theorem 1.5 proves the cyclic sum conjecture, giving the explicit evaluation $(-1)^t N^{-1}\binom{N+t}{r+1}(1-\zeta_N)^r$.
Corollary 1.6 follows: the symmetric cyclic sum equals $(-1)^t(-2\pi i)^r/(r+1)!$ and the finite cyclic sum vanishes.
The argument shows that harmonic-product identities for $q$-series can be transported from formal power series to roots of unity through the finite-sum formulation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Beyond the paper, the coefficient-comparison mechanism is algebraic, so the same proof likely yields further linear relations among finite multiple harmonic $q$-series by varying the chosen basis or the specialization of $t$.
Beyond the paper, because Theorem 3.3 already covers arbitrary tuples in $F^+$ (indices with all entries at least $2$), the method is broader than the block patterns of Conjecture 1.4 and could produce additional cyclic sum identities not written down in the paper.
Beyond the paper, one could test twisted versions by replacing $\zeta_N$ with $\zeta_N^a$ for $a$ coprime to $N$; the continuity step should survive, giving a family of root-of-unity cyclic sum formulas.
Beyond the paper, the boundary-continuation step might be reproved by direct estimates on the defining sums, which would make the analytic hypothesis explicit and could extend the theorem to other $q$ on the unit circle.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The paper proves an analogue of Hirose's relation for finite multiple harmonic q-series at roots of unity and, as an application, proves the cyclic sum conjecture of Hessami Pilehrood, Hessami Pilehrood, and Tauraso. That is a genuine result: the conjecture was open, and the main theorem is a real q-analogue, not a trivial deformation of Hirose's theorem. The method is the connected-sum technique of Seki and Yamamoto, extended to the q-setting; the argument is mostly algebraic and verifiable.

What the paper does well: Theorem 1.2 is the finite/cyclotomic counterpart of Hirose's theorem, Theorem 2.5 proves an equivalent version, and Theorem 1.5 deduces the cyclic sum formula from a Tanaka–Wakabayashi type identity. The proof of Lemma 2.4 uses Heine's summation correctly, and the coefficient comparison in the proof of Theorem 2.5 is legitimate because the family forms a topological basis. The paper also gives a new proof of Hirose's theorem by taking N→∞ and a finite analogue by reducing modulo p; these are nice bonuses but not the core.

The main soft spot is the continuity extension of the duality formula to boundary specializations s=1 or t=1. The paper asserts this in a sentence; I checked it and it works: for admissible k the relevant series converges absolutely on |q|≤r<1, |s|,|t|≤1, so dominated convergence justifies the boundary. The later limit q→ζ_N is taken in a finite rational coefficient identity after multiplying by (1−q^N)^2, so no pole issue arises. The equivalence between Theorem 1.2 and 2.5 is also terse but correct: the extra (1−q) term in ∗_q is handled by the normalization (1−q)^{−wt}.

Minor presentational issues: the notation is dense, and the paper could state the continuity argument in more detail. These are not substantive.

Who it's for: anyone working on finite multiple zeta values, symmetric MZVs, or q-analogues at roots of unity. It's a specialized but solid contribution. It deserves refereeing; a careful referee should ask for a fuller justification of the boundary extension, but that is a minor revision, not a block.

Referee Report

0 major / 5 minor

Summary. The paper establishes a finite analogue of Hirose's relation for finite multiple harmonic q-series at roots of unity. The main result, Theorem 1.2, states that for non-empty indices k and l whose first and last entries are at least 2, one has z_N((↓(k *_{ζ_N} l)↓)^∨) = 0. This is proved via the connected-sum method of Seki and Yamamoto, using a duality formula for q-series and a topological basis coefficient comparison. The paper then derives Theorem 1.5, proving the cyclic sum conjecture of Hessami Pilehrood, Hessami Pilehrood, and Tauraso, by combining Theorem 2.5 with the Tanaka–Wakabayashi algebraic cyclic sum identity and special values from Bachmann–Takeyama–Tasaka. Letting N tend to infinity recovers Hirose's theorem, and reducing modulo p gives the corresponding finite multiple zeta value statement.

Significance. If correct, the result gives the first finite analogue of Hirose's relation and settles the cyclic sum conjecture for finite multiple harmonic q-series at roots of unity. A notable strength is that the proof is self-contained relative to standard external inputs: Heine's summation, the connected-sum duality, the Tanaka–Wakabayashi quasi-shuffle identity, and the special-value theorem of Bachmann–Takeyama–Tasaka. The argument does not assume the cyclic sum conjecture or Hirose's theorem, and no fitted parameters or circular steps are apparent. The paper also offers a new proof of Hirose's theorem in the N→∞ limit and yields both symmetric and finite corollaries.

minor comments (5)

[§2.2] The sentence 'Theorem 1.2 can be rewritten as follows' skips the verification that replacing the q-harmonic product ∗_q with ∗_+ changes z_N only by a nonzero factor. Please add the explicit computation showing z_N((↓(k∗_q l)↓)^∨) = (1−q)^{wt(k)+wt(l)−2} \bar z_N((↓(k∗_+ l)↓)^∨).
[§2.1, Lemma 2.4] The summation in Lemma 2.4 is written as \sum_{n+m≤r} with r free; it should be \sum_{r≥n+m}. The proof also leaves the summation over r implicit in the displayed formula, which makes the subsequent basis comparison harder to follow.
[§2.2] The assertion that the defining series ζ_q(k;s,t) converges absolutely at the boundary s=1 or t=1 for admissible k is stated without proof. For admissible indices the denominators (1−s q^{n_i}) and (tq;q)_{n_a} are bounded away from zero on |q|≤r<1 and |s|,|t|≤1, so dominated convergence justifies the extension; a brief sentence to this effect would improve readability.
[§3.3] In the proof of Theorem 3.3 the identity f(j) = z_N({1}^{j−2}) is used; it may help to state explicitly that z_N of the empty index is taken to be 1 when j=2, so that the displayed formula is consistent with the boundary case.
[References] The reference [KZ] is listed as 'to appear'; if a published version now exists, the citation should be updated.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the finite q-analogue of Hirose's relation is derived from the connected-sum machinery and external lemmas (BTT2, TW, SY, GR), not from the conjecture or from any fitted or author-supplied input.

full rationale

The paper's central derivation is self-contained. Theorem 1.2 is proved by first rewriting it as Theorem 2.5, which is a legitimate equivalence (both statements define the same normalized finite sums z_N with the q-harmonic product specialized to roots of unity), then proving Theorem 2.5 from Proposition 2.2 (duality via transport relations), Corollary 2.3, Lemma 2.4 (obtained from Heine's summation formula), and a basis argument in C[[q]][[t]]. The limit q→ζ_N is taken only in a finite rational identity after multiplying by (1−q^N)^2, with the explicit claim that the only pole is removed; the boundary extension of Proposition 2.2 to s=1 or t=1 is asserted by absolute convergence, but this is an analytic justification, not a circular one. Theorem 1.5 uses an externally proved algebraic identity of Tanaka–Wakabayashi (Lemma 3.1), a variant proved by the same recursion, and the independent special-value computation [BTT2, Theorem 1.1] for f(j)=z_N({1}^{j−2}). There are no fitted parameters, no author-generated inputs, no self-citation chain, and no claim that is equivalent to Conjecture 1.4 by construction; the conjecture is proved, not assumed. Even the weakest analytical step (continuity extension) is not circular; at most it is a correctness risk. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

The result is a proof-based theorem with no free parameters, no new postulated entities, and no data fitting. It relies on standard convergence facts, the connected-sum duality, known special values from BTT2, and the Tanaka-Wakabayashi algebraic congruence.

assumptions (3)

domain assumption Absolute convergence of ζ_q(k;s,t) at boundary specializations s=1 or t=1 for admissible k
Section 2.2 invokes this to extend Proposition 2.2 to the boundary by continuity before limiting q to a root of unity.
standard math The family ((q;q)_r/(tq;q)_r) t^r for r≥0 is a topological C[[q]]-basis of C[[q]][[t]]
Used in Theorem 2.5 to compare coefficients; the argument is triangularity plus (q;q)_r being a unit in C[[q]].
domain assumption Tanaka-Wakabayashi's cyclic sum congruence remains valid after replacing the letter product by (r)∘_+(s)=(r+s)+(r+s-1)
Assumed in Lemma 3.2; the paper says the proof transfer is justified by commutativity and associativity of the modified product.

how reviews work

0 comments

Cite this review

Pith. "Pith review of An analogue of Hirose's relation for finite multiple harmonic $q$-series at roots of unity." pith.science (2026). https://pith.science/paper/RJFWISTK

@misc{pith2026260809806,
  author       = {Pith},
  title        = {Pith review of: An analogue of Hirose's relation for finite multiple harmonic $q$-series at roots of unity},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/RJFWISTK}},
  note         = {Machine review of arXiv:2608.09806}
}

abstract

Recently, Hirose proved an analogue of the linear part of Kawashima's relations for refined symmetric multiple zeta values. In this paper, we establish an analogue of Hirose's relation for finite multiple harmonic $q$-series at roots of unity. As an application, we prove the cyclic sum conjecture proposed by Kh. Hessami Pilehrood, T. Hessami Pilehrood, and R. Tauraso.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

12 extracted references · 11 canonical work pages

[1]

Bouillot, The algebra of multitangent functions, J

O. Bouillot, The algebra of multitangent functions, J. Algebra 410 (2014), 148--238

work page 2014
[2]

Bachmann, Y

H. Bachmann, Y. Takeyama, K. Tasaka, Cyclotomic analogues of finite multiple zeta values, Compos. Math. 154 (2018), no. 12, 2701--2721

work page 2018
[3]

Bachmann, Y

H. Bachmann, Y. Takeyama, K. Tasaka, Special values of finite multiple harmonic $q$-series at roots of unity, in Algebraic Combinatorics, Resurgence, Moulds and Applications (CARMA), Vol. 2, IRMA Lect. Math. Theor. Phys. 32 (2020), 1--18

work page 2020
[4]

Gasper, M

G. Gasper, M. Rahman, Basic Hypergeometric Series, Encyclopedia of Mathematics and its Applications, vol. 96, 2nd ed., Cambridge University Press, 2004

work page 2004
[5]

Hirose, Double shuffle relations for refined symmetric multiple zeta values, Doc

M. Hirose, Double shuffle relations for refined symmetric multiple zeta values, Doc. Math. 25 (2020), 365--380

work page 2020
[6]

Hirose, Multitangent functions and symmetric multiple zeta values, preprint, 2024, arXiv:2402.13902

M. Hirose, Multitangent functions and symmetric multiple zeta values, preprint, 2024, arXiv:2402.13902

arXiv 2024
[7]

Hessami Pilehrood, T

Kh. Hessami Pilehrood, T. Hessami Pilehrood, R. Tauraso, On 3-2-1 values of finite multiple harmonic q -series at roots of unity, J. Math. Soc. Japan 74 (2022), no. 3, 753--758

work page 2022
[8]

Kawashima, A class of relations among multiple zeta values, J

G. Kawashima, A class of relations among multiple zeta values, J. Number Theory 129 (2009), no. 4, 755--788

work page 2009

Show all 12 references

[9]

Kaneko, D

M. Kaneko, D. Zagier, Finite multiple zeta values, to appear in the Proceedings of the 17th MSJ-SI conference on Modular forms and Multiple Zeta values
[10]

Seki, Connectors, RIMS K\^oky\^uroku, 2160 (2020), 15--27

S. Seki, Connectors, RIMS K\^oky\^uroku, 2160 (2020), 15--27

2020
[11]

S. Seki, S. Yamamoto, A new proof of the duality of multiple zeta values and its generalizations, Int. J. Number Theory 15 (2019), no. 6, 1261--1265

2019
[12]

Tanaka, N

T. Tanaka, N. Wakabayashi, An algebraic proof of the cyclic sum formula for multiple zeta values, J. Algebra 323 (2010), no. 3, 766--778

2010

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Bouillot, The algebra of multitangent functions, J

O. Bouillot, The algebra of multitangent functions, J. Algebra 410 (2014), 148--238

work page 2014

[2] [2]

Bachmann, Y

H. Bachmann, Y. Takeyama, K. Tasaka, Cyclotomic analogues of finite multiple zeta values, Compos. Math. 154 (2018), no. 12, 2701--2721

work page 2018

[3] [3]

Bachmann, Y

H. Bachmann, Y. Takeyama, K. Tasaka, Special values of finite multiple harmonic \(q\)-series at roots of unity, in Algebraic Combinatorics, Resurgence, Moulds and Applications (CARMA), Vol. 2, IRMA Lect. Math. Theor. Phys. 32 (2020), 1--18

work page 2020

[4] [4]

Gasper, M

G. Gasper, M. Rahman, Basic Hypergeometric Series, Encyclopedia of Mathematics and its Applications, vol. 96, 2nd ed., Cambridge University Press, 2004

work page 2004

[5] [5]

Hirose, Double shuffle relations for refined symmetric multiple zeta values, Doc

M. Hirose, Double shuffle relations for refined symmetric multiple zeta values, Doc. Math. 25 (2020), 365--380

work page 2020

[6] [6]

Hirose, Multitangent functions and symmetric multiple zeta values, preprint, 2024, arXiv:2402.13902

M. Hirose, Multitangent functions and symmetric multiple zeta values, preprint, 2024, arXiv:2402.13902

arXiv 2024

[7] [7]

Hessami Pilehrood, T

Kh. Hessami Pilehrood, T. Hessami Pilehrood, R. Tauraso, On 3-2-1 values of finite multiple harmonic q -series at roots of unity, J. Math. Soc. Japan 74 (2022), no. 3, 753--758

work page 2022

[8] [8]

Kawashima, A class of relations among multiple zeta values, J

G. Kawashima, A class of relations among multiple zeta values, J. Number Theory 129 (2009), no. 4, 755--788

work page 2009