Comment on "Chiral symmetry restoration, the eigenvalue density of the Dirac operator, and the axial U(1) anomaly at finite temperature"

Matteo Giordano

REVIEW 3 major objections 4 minor 10 references

Comment on "Chiral symmetry restoration, the eigenvalue density of the Dirac operator, and the axial U(1) anomaly at finite temperature"

T0 review · 3 major / 4 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A crucial inference in a proof that chiral symmetry forces the Dirac spectral density and topological susceptibility to vanish at small quark mass is invalid: one power of a positive gauge-field observable vanishing like $m^{k_0}$ does…

desk verdict A real gap in AFT's moment-inference step, shown with clean new counterexamples, but the comment never confirms the actual l0 in their proof, so the refutation of their specific conclusion is incomplete. read the letter →

arxiv 2510.24403 v3 pith:4CK24UPL submitted 2025-10-28 hep-lat hep-phhep-th

Matteo Giordano This is my paper · ORCID

classification hep-lathep-phhep-th

keywords chiralsymmetryrestorationDiracoperatorspectraldensitytopologicalsusceptibilityaxialU(1)anomalyfinite-temperatureQCDofstateslatticemassdependenceobservables

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 comment targets a specific step in a proof that, in the chirally symmetric phase of two-flavor QCD, the Dirac spectral density at the origin and the topological susceptibility must vanish identically for sufficiently small nonzero quark mass $m$. The step claims that if one positive gauge-field observable satisfies $\langle O(A)^{l_0}\rangle=O(m^{k_0})$ as $m\to 0$, then every power satisfies $\langle O(A)^l\rangle=O(m^{k_0})$ for all $l\ge 1$. The comment shows, through density-of-states counterexamples, that this implication does not follow from positivity or shared support alone: a moment can vanish quickly because the value of $O$ on typical configurations shrinks, rather than because the contributing region of configuration space shrinks. If so, the original proof's conclusions $\rho(0;m)=0$ and $\chi_t=0$ require a missing argument. An appended note also refutes a reply's objection by showing that the assumptions invoked would force all local gluonic correlators to be independent of $m$ at small mass, which QCD does not exhibit.

What carries the argument

The density-of-states function $p_O(x,m)=\langle\delta(x-O(A))\rangle$ for a positive gauge-field observable $O(A)$ carries the argument. It separates two ways a moment can vanish: the support of $O$-values can shrink toward $x=0$ (as in $p_O=\delta(x-x_0(m))$, where $\langle O^l\rangle=x_0(m)^l$), or the support of gauge-field configurations contributing to the expectation can shrink in measure while $O$ itself stays bounded, which is the original proof's hidden assumption. The counterexample distributions—a single delta moving to zero, a truncated Gaussian, and any density with a scaling form $p_O(x,m)=x_0(m)^{-1}\pi_O(x/x_0(m),0)$—show that one moment's vanishing rate does not control the others. The key comparison is $\langle O^l\rangle/\langle O^{l_0}\rangle\propto x_0^{l-l_0}$ or $\sigma^{l-l_0}$, which blows up for $l<l_0$ whenever the relevant scale vanishes.

What would settle it

On a lattice with two light quark flavors in the chirally symmetric phase, compute $p_O(x,m)=\langle\delta(x-O(A))\rangle$ and the moments $\langle O^l\rangle$ for a positive local gluonic observable $O$ at several small masses $m$. If $p_O(x,m)$ has a peak whose position $x_0(m)$ vanishes as $m\to 0$ while its width is comparable to $x_0(m)$, or more generally if the rescaled density $x_0^{-1}p_O(x/x_0,m)$ has a nonzero chiral limit, then the ratios $\langle O^l\rangle/\langle O^{l_0}\rangle$ will not stay of order one for different $l$, demonstrating that Eq. (1) does not imply Eq. (3).

Watch

Extended reading notes

Core claim

The central claim, stated on the paper's own terms, is that Eq. (3) is not a necessary consequence of Eq. (1). Writing $\langle O(A)^l\rangle=\int dx\,x^l p_O(x,m)$ with $p_O(x,m)=\langle\delta(x-O(A))\rangle$, the inference would require $p_O$ to have the special form $[1-\alpha(m)]\delta(x)+\alpha(m)\tilde p(x;m)$ with $\alpha(m)=O(m^{k_0})$. But ordinary distributions invalidate this: if $p_O(x,m)=\delta(x-x_0(m))$ with $x_0(m)\to 0$, then $\langle O^l\rangle=x_0(m)^l$, so a single moment of order $l_0$ being $O(m^{k_0})$ does not make lower moments $O(m^{k_0})$ unless $l_0=l$. A truncated Gaussian with mean and width vanishing in the chiral limit behaves similarly; if the width-to-mean ratio has a nonzero or infinite chiral limit, ratios such as $\langle O^l\rangle/\langle O^{l_0}\rangle$ blow up for $l<l_0$. Thus the load-bearing inference in the original proof fails, and the claimed vanishing of $\rho(0;m)$ and $\chi_t$ for small nonzero $m$ is not established by that argument.

Load-bearing premise

The counterexamples presume that the infinite-volume QCD expectation value has a well-defined measure on gauge-field configurations and that density-of-states functions behave like ordinary probability densities; if those measures do not exist or cannot be realized in QCD, the logical gap may remain real but the explicit counterexamples would not carry over.

Editorial extensions

If this is right

The specific proof in the original article no longer establishes $\rho(0;m)=0$ or $\chi_t=0$ for small nonzero $m$; these conclusions need a new argument that supplies the missing inference from one moment to all moments.
The claimed necessity of a phase transition at small nonzero quark mass in the high-temperature phase, drawn from $\chi_t=0$, loses its proof-based support and remains an open numerical question.
The fate of the axial U(1) anomaly in the chiral limit is not settled by this proof route; the scalar-pseudoscalar sector analysis in the original article depends on the same unproven step.
Future proofs of the same conclusion must either impose a density-of-states structure (a fixed delta peak plus a smooth part with bounded moments) or otherwise control all moments, not just one.
The appended note adds a consistency constraint: any argument that both assumes $m^2$-analyticity of gluonic observables and uses the disputed implication would predict that all local gluonic correlators are mass-independent at small $m$, which contradicts observed QCD behavior.

Reading between the lines

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

The counterexample construction transfers to any positive observable whose expectation value is controlled by a self-averaging value rather than by the size of the support; similar 'one moment controls all moments' steps in other spectral or thermodynamic arguments should be checked against the same density-of-states test.
A concrete lattice test suggested by the paper's own language: measure $p_O(x,m)$ for a simple positive gluonic quantity in the high-temperature phase; if the distribution shifts toward zero with a width comparable to its mean, the moments will scale with distinct powers of $m$, directly demonstrating the failure mode.
If the density-of-states distributions are not ordinary functions on the infinite-volume configuration space, the counterexamples may not transfer verbatim to QCD, but the burden is then on the original proof to justify the measure-theoretic setup it assumes; the logical gap itself is independent of that technical caveat.
The same type of argument could be tested in zero-dimensional or random-matrix models where the density of states is exactly known, providing a cheap check of whether the missing inference holds in simplified settings.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

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

Dear colleague,

The thing to know: this is a genuine, well-aimed shot at one step in Aoki-Fukaya-Taniguchi 2012. Giordano shows the inference from = O(m^{k0}) to = O(m^{k0}) for all l is not a theorem. The truncated-Gaussian and delta-function density-of-states counterexamples are explicit, simple, and, as far as I know, new. That alone earns the paper a serious look.

What is good: the comment does not just complain. It isolates Eqs. (23)-(25) of AFT, recasts the missing step in density-of-states language, and constructs distributions where one moment vanishes like m^{k0} while lower moments vanish more slowly. The point that positivity and shared support do not force the same-order suppression is correct, and the examples make it undeniable. The author also openly labels his measure construction heuristic, and the citation list covers both sides of the lattice controversy without obvious bias.

Where it is soft: the transfer to AFT's proof is not actually completed. Every counterexample involves l < l0. If AFT only use the lemma with l0 = 1, then for a bounded positive observable the conclusion for all l >= 1 follows trivially via O^l <= C^{l-1} O. Giordano asserts in Sec. G that the implication is crucially used after Eq. (92), but never states what l0 is there. Until he shows that AFT's application requires l0 > 1, the comment has a hole at exactly the load-bearing joint. Second, the abstract promises an additional note responding to Aoki-Fukaya's reply, and that note is absent from this version; that is a real completeness problem. Third, the infinite-volume measures and density-of-states functions are heuristic, so the counterexamples are mathematical analogies unless one argues they are realizable in QCD.

Net assessment: the general lemma is dead; whether AFT's specific conclusion falls is still open. This is a legitimate comment that deserves peer review, but the referee should send it back for the author to verify the actual l0 in AFT's Eq. (92) neighborhood and to supply the promised note. If l0 turns out to be 1, the comment becomes a useful warning about a non-existent general proof, not a refutation of AFT's result.

Best,

Referee Report

3 major / 4 minor

Summary. The manuscript is a Comment on Aoki, Fukaya, and Taniguchi (AFT, Phys. Rev. D 86, 114512), which claims that in the chirally symmetric phase of two-flavor QCD the Dirac spectral density at the origin and the topological susceptibility vanish identically for sufficiently small nonzero quark mass. The Comment isolates a step in AFT's proof in which vanishing of the expectation value of O(A)^{l0} for a positive gauge-field observable O and some l0 is used to infer vanishing of <O(A)^l> for all l. The author argues that Eq. (2) does not follow from Eq. (1), that Eq. (3) does not follow generally from Eq. (2), and supports this with a heuristic measure-support construction in Sec. d and with counterexamples based on truncated Gaussian and delta density-of-states functions in Sec. f. The paper concludes that AFT's results for rho(0;m) and chi_t should be reassessed. The abstract also promises an additional note refuting objections raised by Aoki and Fukaya in their reply, but no such note appears in the submitted text.

Significance. If the objection lands, the Comment identifies a genuine logical gap in an influential proof and would change the evidentiary status of the AFT conclusion that U(1)_A is effectively restored in the chiral limit at high temperature. The paper is valuable for making the l-dependence issue explicit, and it provides clean analytic counterexamples to the general moment-inference lemma. It is also free of circularity and does not smuggle the target conclusion into its assumptions. The force of the Comment, however, is conditional on showing that the contested inference is actually used in AFT with l0 > 1, and on supplying the promised additional note; both points are currently unresolved.

major comments (3)

[Sec. g, paragraph after Eq. (92)] The Comment asserts that the implication (1)->(3) is 'crucially used' after Eq. (92) of Ref. [1], but it never identifies the exponent l0 in that application. The counterexamples in Sec. f produce a failure only for powers l < l0: for instance, Eq. (8) gives <O^l>/<O^{l0}> proportional to x0(m)^{l-l0} or sigma(m)^{l-l0}, which blows up only when l < l0. If the actual application in AFT uses l0 = 1, then for a bounded positive observable O the inference is trivially valid because <O^l> <= C^{l-1}<O> for all l >= 1, and the counterexamples are irrelevant. The manuscript must quote the relevant observable and l0 from Ref. [1] and demonstrate that l0 >= 2; otherwise the central conclusion that AFT's proof collapses is not established.
[Abstract, final two sentences] The abstract promises an additional note refuting the objections raised by S. Aoki and H. Fukaya in their reply, and asserts that their claim of having found a mistake in the author's arguments is baseless. No such note appears in the submitted text; the paper ends with Sec. g and the acknowledgments. This is an unfulfilled substantive promise. Either the additional note must be included, or the abstract must be amended to remove the unsupported claim.
[Sec. f, Eqs. (7)-(9)] The counterexamples are constructed from arbitrary probability densities p_O(x,m) that are assumed to be the density-of-states functions of a positive gauge-field observable under the infinite-volume QCD measure. Section d explicitly waives 'any issue of rigor in the construction of the various measures involved'. No argument is given that a truncated Gaussian or a delta distribution can actually arise as p_O for any local gauge-invariant observable in the QCD ensemble, nor that the required infinite-volume measures exist. As written, the examples invalidate the general moment-inference lemma for abstract random variables, but they do not by themselves show that the specific class of observables and assumptions used in Ref. [1] are inconsistent with Eq. (3). The author should either supply a realizability argument or explicitly frame the conclusion as shifting the burden of proof rather than as a full refutation.

minor comments (4)

[Sec. b, Eq. (3)] The text says Eq. (3) holds for 'arbitrary non-negative integer l', but for l = 0 the statement is false because <O^0> = 1. The later text in Sec. g correctly restricts to l > 0; Eq. (3) and its surrounding discussion should be corrected accordingly.
[Sec. d] The symbol P(m,A) is reused for the original non-negative quantity and for the rescaled quantity m^{-k0} P(m,A) chi_{S(O)}(A). Please introduce a distinct symbol, for example ilde P(m,A), to avoid confusion in Eqs. (2)-(4).
[Sec. d] The notation chi_{S_m(P)}(A) is not defined; the support set of the rescaled measure should be defined explicitly or written out in words.
[Sec. b, Eq. (1)] The phrase 'for some non-negative integers l0, k' is ambiguous because k0 is introduced only later, and the relation k0 > k is stated after the definitions. It would be clearer to define k0 before Eq. (2) and state the inequality explicitly there.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the comment builds self-contained counterexamples to an external proof step rather than deriving its conclusion from its own inputs.

full rationale

This is a comment criticizing the logic of Aoki-Fukaya-Taniguchi (AFT), and it does not exhibit any of the circularity patterns defined for this pass. The paper's central claim is that AFT's step from Eq. (1) to Eq. (3) is not justified, and the support for that claim is a set of explicit counterexamples constructed from density-of-states functions: the truncated Gaussian distribution in Eq. (7) and the delta-function example satisfy the premise for some l0 while violating the conclusion for l < l0. These counterexamples are not fitted to data and are not justified by any self-citation; they are mathematically self-contained within the stated assumptions about the density-of-states formalism. The comment even labels its construction heuristic and explicitly ignores issues of rigor in the infinite-volume measures, which is a limitation of the argument but not a circular step. The only external references are lattice studies cited as context for the unsettled status of AFT's predictions, so the citations are not load-bearing for the logical counterexample. The skeptical concern that the counterexamples may not engage the specific l0 used in AFT's Sec. III G is a legitimate correctness question about whether the attack applies to AFT's particular proof, but it is not a circularity. No equation is defined in terms of the conclusion, no fitted quantity is renamed a prediction, and no self-citation chain forces the result. Accordingly, the appropriate finding is no significant circularity, with score 0.

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

The comment is a logical critique and introduces no fitted parameters or new physical entities. It relies on the standard QCD path-integral framework plus explicit, flagged heuristic assumptions about infinite-volume measures and density-of-states functions.

assumptions (3)

domain assumption The infinite-volume expectation value can be represented as an integral over gauge configurations with a measure DAP(m,A) that is well-behaved in the chiral limit.
Section d: 'One starts by writing the thermodynamic limit... for a suitable measure... assumed to be well-behaved in the chiral limit.' This is needed for the heuristic derivation and for the counterexamples to transfer to QCD.
domain assumption The thermodynamic limit is taken before the chiral limit in the expectation values in Eqs. (1)-(3).
Section c states that for Eq. (1) to be a meaningful statement about the chiral limit, the thermodynamic limit should be taken first. This ordering is a physical assumption about how lattice limits are taken.
standard math The density-of-states representation pO(x,m)=<delta(x-O(A))> exists as a probability measure on x>=0 for positive observables.
Section f introduces this representation as a formal but standard tool. Existence of such distributions in infinite volume is assumed rather than proved.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Comment on "Chiral symmetry restoration, the eigenvalue density of the Dirac operator, and the axial U(1) anomaly at finite temperature"." pith.science (2026). https://pith.science/paper/4CK24UPL

@misc{pith2026251024403,
  author       = {Pith},
  title        = {Pith review of: Comment on "Chiral symmetry restoration, the eigenvalue density of the Dirac operator, and the axial U(1) anomaly at finite temperature"},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4CK24UPL}},
  note         = {Machine review of arXiv:2510.24403}
}

abstract

Aoki, Fukaya, and Taniguchi claim that both the spectral density of the Dirac operator at the origin and the topological susceptibility must vanish identically for sufficiently small but nonzero quark mass $m$ in the chirally symmetric phase of quantum chromodynamics with two light quark flavors, under certain technical assumptions on the spectrum and on the dependence of observables on $m$. I argue that a crucial step of their proof is not justified, and the validity of these conclusions should be reassessed. In an additional note I refute the objections raised by S. Aoki and H. Fukaya in their reply to my comment. I show that both the $m^2$-analyticity assumption they invoke against my arguments, and the very claim that these arguments criticize, imply that all local gluonic correlators are independent of the light-quark mass at small mass. This behavior, however, is not observed in quantum chromodynamics. I then show that their claim of having found a mistake in my arguments is baseless.

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

10 extracted references · 10 linked inside Pith

[1]

S. Aoki, H. Fukaya, and Y. Taniguchi, Phys. Rev. D86, 114512 (2012), arXiv:1209.2061 [hep-lat]

arXiv 2012
[2]

Cossu, S

G. Cossu, S. Aoki, H. Fukaya, S. Hashimoto, T. Kaneko, H. Matsufuru, and J.-I. Noaki, Phys. Rev. D87, 114514 (2013), [Erratum: Phys. Rev. D 88, 019901 (2013)], arXiv:1304.6145 [hep-lat]

arXiv 2013
[3]

Tomiya, G

A. Tomiya, G. Cossu, S. Aoki, H. Fukaya, S. Hashimoto, T. Kaneko, and J. Noaki, Phys. Rev. D96, 034509 (2017), [Addendum: Phys. Rev. D 96, 079902 (2017)], arXiv:1612.01908 [hep-lat]

arXiv 2017
[4]

S. Aoki, Y. Aoki, G. Cossu, H. Fukaya, S. Hashimoto, T. Kaneko, C. Rohrhofer, and K. Suzuki (JLQCD collaboration), Phys. Rev. D103, 074506 (2021), arXiv:2011.01499 [hep-lat]

arXiv 2021
[5]

S. Aoki, Y. Aoki, H. Fukaya, S. Hashimoto, I. Kanamori, T. Kaneko, Y. Nakamura, C. Rohrhofer, K. Suzuki, and D. Ward (JLQCD collaboration), PoSLA TTICE2023, 185 (2024), arXiv:2401.14022 [hep-lat]

arXiv 2024
[6]

Chandrasekharan, D

S. Chandrasekharan, D. Chen, N. H. Christ, W.-J. Lee, R. Mawhinney, and P. M. Vranas, Phys. Rev. Lett.82, 2463 (1999), arXiv:hep-lat/9807018

arXiv 1999
[7]

V. Dick, F. Karsch, E. Laermann, S. Mukherjee, and S. Sharma, Phys. Rev. D91, 094504 (2015), arXiv:1502.06190 [hep-lat]

arXiv 2015
[8]

Dinget al.(HotQCD collaboration), Phys

H.-T. Dinget al.(HotQCD collaboration), Phys. Rev. Lett.123, 062002 (2019), arXiv:1903.04801 [hep-lat]

arXiv 2019

Show all 10 references

[9]

H. T. Ding, S. T. Li, S. Mukherjee, A. Tomiya, X. D. Wang, and Y. Zhang, Phys. Rev. Lett.126, 082001 (2021), arXiv:2010.14836 [hep-lat]

2021 arXiv
[10]

Kaczmarek, L

O. Kaczmarek, L. Mazur, and S. Sharma, Phys. Rev. D 104, 094518 (2021), arXiv:2102.06136 [hep-lat]

2021 arXiv

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [1]

S. Aoki, H. Fukaya, and Y. Taniguchi, Phys. Rev. D86, 114512 (2012), arXiv:1209.2061 [hep-lat]

arXiv 2012

[2] [2]

Cossu, S

G. Cossu, S. Aoki, H. Fukaya, S. Hashimoto, T. Kaneko, H. Matsufuru, and J.-I. Noaki, Phys. Rev. D87, 114514 (2013), [Erratum: Phys. Rev. D 88, 019901 (2013)], arXiv:1304.6145 [hep-lat]

arXiv 2013

[3] [3]

Tomiya, G

A. Tomiya, G. Cossu, S. Aoki, H. Fukaya, S. Hashimoto, T. Kaneko, and J. Noaki, Phys. Rev. D96, 034509 (2017), [Addendum: Phys. Rev. D 96, 079902 (2017)], arXiv:1612.01908 [hep-lat]

arXiv 2017

[4] [4]

S. Aoki, Y. Aoki, G. Cossu, H. Fukaya, S. Hashimoto, T. Kaneko, C. Rohrhofer, and K. Suzuki (JLQCD collaboration), Phys. Rev. D103, 074506 (2021), arXiv:2011.01499 [hep-lat]

arXiv 2021

[5] [5]

S. Aoki, Y. Aoki, H. Fukaya, S. Hashimoto, I. Kanamori, T. Kaneko, Y. Nakamura, C. Rohrhofer, K. Suzuki, and D. Ward (JLQCD collaboration), PoSLA TTICE2023, 185 (2024), arXiv:2401.14022 [hep-lat]

arXiv 2024

[6] [6]

Chandrasekharan, D

S. Chandrasekharan, D. Chen, N. H. Christ, W.-J. Lee, R. Mawhinney, and P. M. Vranas, Phys. Rev. Lett.82, 2463 (1999), arXiv:hep-lat/9807018

arXiv 1999

[7] [7]

V. Dick, F. Karsch, E. Laermann, S. Mukherjee, and S. Sharma, Phys. Rev. D91, 094504 (2015), arXiv:1502.06190 [hep-lat]

arXiv 2015

[8] [8]

Dinget al.(HotQCD collaboration), Phys

H.-T. Dinget al.(HotQCD collaboration), Phys. Rev. Lett.123, 062002 (2019), arXiv:1903.04801 [hep-lat]

arXiv 2019