{"id":"65f14424-8cdb-4b9a-a98c-335571ec628a","arxiv_id":"1908.01676","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"The rounded clover topological charge at large Wilson flow time gives the same topological susceptibility as the overlap-Dirac index, despite per-configuration differences on 31.2% of configurations.","lead":"Using 535 lattice QCD configurations, the authors compared two ways of measuring topological charge: the exact overlap-Dirac index and a rounded clover charge after Wilson flow. The two methods disagree on 31.2% of individual configurations, yet they produce the same topological susceptibility.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Unreconciled 31.2% per-configuration difference between Qtc and index(D_o) means the paper validates only the second moment; its claim that Qtc captures χt, c4, and all topological fluctuations is not established.","rationale":"Good-faith reading: the numerical work is internally consistent and has real independent support — exact overlap indices on each configuration, a well-defined Wilson-flow integration, and the exact equality of Qtc with the overlap index computed on the flowed configuration at t=77. The reader's CONDITIONAL verdict is therefore appropriate. The load-bearing assumption is that Qtc equals the genuine topological charge of the underlying gauge background. The paper's evidence for this is the agreement of the second moment of the charge distribution, while 31.2% of configurations disagree per configuration. Since the final sentence claims that χt, c4, ... can be obtained from Qtc, and no higher-moment test is given, the concern is concrete and central rather than a matter of taste. The authors themselves flag the limitation for non-chiral fermions and for coarser lattices. My check would clarify whether the agreement extends beyond ⟨Q²⟩; absent that check, the conditional verdict stands unchanged.","tokens_in":7340,"tokens_out":4987,"duration_ms":57123,"concrete_test":"Perform a goodness-of-fit test between the full histograms of Qtc and index(D_o) over all 535 configurations, specifically comparing the fourth cumulant c4 = (⟨Q^4⟩ − 3⟨Q^2⟩^2)/V with jackknife errors. If c4 differs by more than the combined statistical error, or if a chi-square or Kolmogorov–Smirnov test rejects distributional equality, then the claim that Qtc captures all topological fluctuations (χt, c4, · · ·) is not supported by this ensemble.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that the asymptotically-invariant Qtc from the Wilson flow is the genuine topological charge, so that χt and higher moments can be computed from it. The quantitative support is the agreement of the second moment: χt a^4 = 6.03(61)×10^-7 (Eq. 6) versus 7.03(91)×10^-7 (Eq. 4), together with the visual statement that the histograms in Fig. 1 are 'almost identical.' This is insufficient because 167 of 535 configurations (31.2%) have Qtc ≠ index(D_o) at t=0. Equality of the second moment does not imply equality of the full distribution; the paper explicitly invokes 'higher moments (c4, · · ·)' in its concluding paragraph but never computes or tests them. The verification that index(D_o) at t=77 equals Qtc is not independent: it only shows that Wilson flow shifts the overlap index for 31.2% of configurations, so the choice of flow time at which 'topology' is defined is itself a convention. The paper's own final paragraph concedes that for non-chiral fermions, or at lattice spacings a > 0.1 fm, it is uncertain whether Qtc captures genuine topological fluctuations. These caveats, combined with the unresolved per-configuration mismatch, make the inference from one second moment on one ensemble to 'topological fluctuations of the QCD vacuum' the load-bearing weak point.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the topological charge in N_f=2 lattice QCD with the optimal domain-wall fermion, using 535 configurations on a 24^4 x 48 lattice at a = 0.06 fm and M_pi = 260 MeV. For each configuration, the index of the overlap-Dirac operator is computed at flow time t=0, and the clover topological charge Q_clover(t) is integrated along the Wilson flow from t=0 to t=128 with delta_t=0.01. The authors observe that the rounded charge Q_c(t) = round[Q_clover(t)] becomes invariant for t >= t_c, with max{t_c} ~ 77. Comparing the asymptotically invariant Q_{tc} with the overlap index at t=0, they find disagreement on 31.2% of configurations, but the histograms are reported as 'almost identical,' and the topological susceptibility computed from Q_{tc} is chi_t a^4 = 6.03(61) x 10^-7, consistent with the overlap-index value 7.03(91) x 10^-7. They also verify that at t=77 the overlap index exactly equals Q_{tc} for all configurations. The paper concludes that the topological susceptibility, and by extension higher moments, of lattice QCD with exact chiral symmetry can be obtained from the asymptotically invariant Q_{tc} in the Wilson flow.","tokens_in":7625,"tokens_out":3904,"duration_ms":40005,"significance":"If the central result holds, the paper provides a practical prescription for extracting the topological susceptibility from Wilson-flowed clover charges, benchmarked against the exact chiral symmetry of the overlap-Dirac operator. The numerical work is careful: eigenmodes are computed with residual 10^-12, the sign function error is below 10^-14, and the flow is integrated with a fine step to large t. The explicit cross-check that the overlap index at t=77 equals Q_{tc} for every configuration is a useful consistency test. The agreement of the second moment, chi_t a^4 = 6.03(61) x 10^-7 versus 7.03(91) x 10^-7, is a nontrivial quantitative result. However, the paper's reach exceeds its evidence: the per-configuration mismatch of 31.2% means the claim reduces to an equality of the second moment only, and the extension to higher moments (c4, ...) in the concluding paragraph is not supported by any computation. The single-ensemble, single-lattice-spacing character of the study is also a limitation that the authors themselves partially acknowledge in the final paragraph.","major_comments":[{"comment":"The concluding statement that the results imply the topological fluctuations '(χt, c4, ···)' can be obtained from Qtc is not supported by the data presented. The paper computes only the second moment, Eqs. (4) and (6), and shows visually in Fig. 1 that two histograms look similar. Since 167 of 535 configurations have Qtc ≠ index(D_o) at t=0, equality of the second moment does not imply equality of the fourth cumulant or higher moments. To make the claimed inference, the authors should either compute c4 (and ideally a test of distributional equality such as a chi-square or Kolmogorov-Smirnov statistic) from the Qtc and overlap-index histograms, or explicitly restrict the conclusion to χt.","section":"Final paragraph and the paragraph after Fig. 5"},{"comment":"The statement that for t ≥ max{t_c} the ensemble 'decomposes into topological sectors, similar to the gauge fields in the continuum theory' is an assumption, not an established fact. The observed per-configuration difference between Qtc and the overlap index at t=0 means that at least one of the two definitions changes under the flow for 31.2% of configurations. The verification that index(D_o) at t=77 equals Qtc shows consistency of the two definitions on the flowed gauge field, but it does not demonstrate that either equals the topological charge of the original configuration at t=0. The paper should state this more carefully: the susceptibility agreement is evidence for a distribution-level match of the second moment, not for per-configuration identification of the topological sector. A quantitative comparison of higher moments would materially strengthen this load-bearing point.","section":"Section 4, paragraph after Fig. 3 and before Eq. (6)"},{"comment":"The agreement between χt a^4 = 6.03(61) × 10^-7 and 7.03(91) × 10^-7 is within about one standard deviation, but no significance level or systematic difference is quoted. Given that the paper's central numerical claim is this agreement, the authors should report the difference and its combined uncertainty, and ideally a p-value, rather than relying only on overlapping error bars and a visual inspection of histograms.","section":"Eqs. (4) and (6) and Fig. 5"}],"minor_comments":[{"comment":"The title contains a typo: 'susceptibilty' should be 'susceptibility', and the line break in 'cl over' should be removed.","section":"Title"},{"comment":"The text 'agress' should be 'agrees'.","section":"Fig. 5 caption / text near Eq. (6)"},{"comment":"The Wilson flow is integrated with δt = 0.01, but no estimate of the systematic error from the finite step size is given. A brief test with a smaller step on a few configurations would clarify the robustness of the plateau and of the threshold max{t_c}.","section":"Numerical integration section"},{"comment":"The vertical axes are labeled 'P(ΔQc=0)' and 'P(>0.978)' without defining what P denotes; the text implies it is a fraction of configurations, but the notation should be made explicit in the figure captions.","section":"Figs. 2 and 3"}],"recommendation":"major_revision","confidential_remarks":"The paper is a short, focused study whose central numerical comparison is defensible, but the concluding generalization to higher moments is a clear overreach relative to the data. The single-ensemble, single-spacing nature of the test and the 31.2% per-configuration mismatch are acknowledged in the text, yet the abstract's final sentence already commits to the broader claim. The manuscript would be acceptable after the conclusions are tightened to the second moment, or after a computation of c4 is added. I do not see evidence of circularity in the susceptibility comparison itself; the flow-time threshold is data-dependent but not tuned to reproduce the overlap-index susceptibility."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this is a technically careful, mostly honest lattice-QCD methods paper, and the central numerical comparison is defensible. What is new is a per-configuration comparison, on one Nf=2 ensemble at a = 0.06 fm and M_pi = 260 MeV, between the overlap-Dirac index at t=0 and the asymptotic Wilson-flow clover charge Qtc, with the explicit threshold max{tc} ~ 77. The authors report that 167 of 535 configurations (31.2%) change integer charge along the flow, yet the histograms and the susceptibility agree: chi_t a^4 = 7.03(91) x 10^-7 (overlap) vs 6.03(61) x 10^-7 (clover). The numerics are documented at a level that makes the result reproducible: residuals of 10^-12 for eigenmodes, sign function error < 10^-14, flow integration to t=128 with delta_t=0.01, and a check that the overlap index at t=77 equals Qtc for every configuration.\n\nWhere it gets soft: the paper's abstract and conclusion say this implies chi_t, c4, and all topological fluctuations can be obtained from Qtc. That is not established. Only the second moment has been shown to agree. Higher moments are not computed, and a 31.2% per-configuration mismatch means the flow is not preserving the sector assignment; the agreement of the second moment could be accidental or ensemble-specific. The authors themselves concede at the end that for non-chiral fermions or a > 0.1 fm it is unknown whether Qtc captures genuine fluctuations, which undercuts the sweeping phrasing. Also, the verification at t=77 only shows that the overlap index follows the flow; it does not independently validate Qtc as the true topological charge.\n\nThe stress-test note is right on target. The paper does exactly one controlled check and then asserts a general conclusion. That said, the result is still useful as a methodological data point: on a fine lattice with exact chiral symmetry, the Wilson-flow clover charge with an asymptotic threshold is a much cheaper observable for chi_t. The free parameter max{tc} is chosen from the data but not tuned to force the agreement, which is a point in its favor.\n\nWho is this for? Lattice practitioners working on topological susceptibility and Wilson-flow smoothing. It deserves a serious referee, not a desk reject. The referee should push for either higher moments (c4) on this ensemble or a more careful statement of what the evidence actually supports.","headline":"A careful lattice-QCD comparison of two topological charge definitions on a single ensemble; the per-configuration mismatch is openly reported, but the paper claims more about higher moments than the evidence supports.","tokens_in":8165,"tokens_out":3130,"would_cite":true,"duration_ms":31451,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The rounded clover topological charge in Wilson flow reproduces the overlap-Dirac index susceptibility in lattice QCD.","keywords":["lattice QCD","topological susceptibility","Wilson flow","overlap-Dirac operator","exact chiral symmetry","domain-wall fermion","clover topological charge","topological sectors"],"falsifier":"On this same ensemble, compute the fourth cumulant $c_4$ of the topological charge distribution from the overlap index and from $Q_{tc}$. If they disagree beyond statistical error, the claim that $Q_{tc}$ captures the full topological fluctuations of the vacuum is false, because $c_4$ is part of the distribution the paper claims to reproduce.","tokens_in":7125,"feed_emoji":"⚛️","tokens_out":14021,"duration_ms":124005,"temperature":0.7,"pith_summary":"The paper asks whether a purely gluonic smoothing procedure—the Wilson flow—can capture the same topological fluctuations of the QCD vacuum that exact chiral symmetry sees through fermion zero modes. On 535 configurations of $N_f=2$ lattice QCD with the optimal domain-wall quark, the rounded clover charge $Q_c(t)=\\mathrm{round}[Q_{\\mathrm{clover}}(t)]$ becomes invariant after enough flow time (at most $t/a^2\\sim77$), and the susceptibility from these asymptotic values, $\\chi_t a^4=6.03(61)\\times10^{-7}$, agrees with the overlap-Dirac index value $7.03(91)\\times10^{-7}$ at $t=0$. The two charges disagree per configuration in 31.2% of cases, but their histograms are almost identical, so the ensemble-level topological fluctuations match. If this holds generally, topological susceptibility—and the $U_A(1)$ and axion-related phenomenology it feeds—can be extracted from flowed clover charges without computing fermion zero modes.","feed_headline":"Flowed clover charge reproduces QCD topological susceptibility","feed_subtitle":"On 535 configurations, the rounded clover charge gives 6.03(61) x 10^-7, matching the overlap-index value 7.03(91) x 10^-7.","key_machinery":"The load-bearing objects are the overlap-Dirac operator index and the Wilson-flowed clover charge. The index, $\\mathrm{index}(D_o)=n_+-n_-$, counts exact zero modes of definite chirality and is tied to the index theorem, so it provides a reference 'genuine' topological charge even on rough lattice fields. The Wilson flow smooths the gauge field by evolving it through $dB_\\mu/dt=D_\\nu G_{\\nu\\mu}$, averaging over a sphere of radius $\\sqrt{8t}$; the clover charge $Q_{\\mathrm{clover}}(t)$ is computed from the flowed links via the field-strength tensor. Rounding gives $Q_c(t)$, and the paper's criterion is that $Q_c$ becomes invariant for $t\\ge t_c$, meaning the configuration has settled into a topological sector. The argument is carried by comparing the two charges: at $t=0$ they differ per configuration in 31.2% of cases, but at $t\\ge t_c$ the index of the flowed configuration equals $Q_c$ exactly, and the ensemble distributions coincide.","core_discovery":"The paper's central result is that the asymptotic rounded clover charge in the Wilson flow, $Q_{tc}$, is a valid proxy for the topological charge of lattice QCD configurations with exact chiral symmetry. Using 535 configurations at $a\\simeq0.06$ fm and $M_\\pi\\simeq260$ MeV, the authors find that $Q_c(t)=\\mathrm{round}[Q_{\\mathrm{clover}}(t)]$ becomes invariant for $t\\ge t_c$, with $\\max\\{t_c\\}\\sim77$, and that for every configuration the overlap-Dirac index computed on the flowed gauge field at $t=77$ equals $Q_{tc}$. Although 167 of the 535 configurations have $Q_{tc}\\ne\\mathrm{index}(D_o)$ at $t=0$, the probability distributions are nearly identical, and the resulting susceptibilities agree: $\\chi_t a^4=6.03(61)\\times10^{-7}$ from $Q_{tc}$ versus $7.03(91)\\times10^{-7}$ from the overlap index. The paper concludes that Wilson-flow clover charges can replace overlap index computations for determining topological fluctuations in exact-chiral-symmetry lattice QCD.","pith_inferences":["The 31.2% per-configuration mismatch shows the flow moves individual configurations across topological sector boundaries; the paper's agreement is statistical, not a statement that each configuration's charge is preserved. A natural follow-up is to check whether the fourth cumulant $c_4$ from $Q_{tc}$ matches the overlap-index value, since matching only $\\chi_t$ is a weaker test.","Repeating the comparison on coarser lattices ($a>0.1$ fm) would test whether $Q_{tc}$ remains faithful; the paper itself flags lattice artifacts there as a possible place where the picture could change.","If the distribution-level agreement persists in larger volumes, Wilson flow could serve as a much cheaper way to map topological fluctuations across the QCD phase diagram, feeding axion-cosmology estimates without fermion zero-mode computations."],"forward_implications":["Topological susceptibility in exact-chiral-symmetry lattice QCD can be obtained from Wilson-flowed clover charges, avoiding the cost of projecting overlap zero modes for every configuration.","The theoretically justified protocol is to flow every configuration until its rounded clover charge is invariant (i.e., up to $t=\\max\\{t_c\\}$), rather than stopping at the susceptibility plateau near $t\\sim10$.","Because the histograms of $Q_{tc}$ and the overlap index are almost identical, the paper asserts that the full topological charge distribution—and higher moments such as $c_4$ and beyond—can also be obtained from $\\{Q_{tc}\\}$.","The overlap-Dirac index is almost invariant under the Wilson flow: computing it at $t=t_c$ reproduces $Q_{tc}$, so the exact-chiral-symmetry charge is stable once configurations become smooth."],"supporting_citations":[{"why":"Defines the Wilson flow and its smoothing properties, and supplies the sufficient condition for a lattice configuration to lie in a topological sector.","marker":"[4]"},{"why":"Defines the overlap-Dirac operator whose index provides the reference genuine topological charge used for comparison.","marker":"[6]"},{"why":"Gives the optimal domain-wall quark action whose effective 4D Dirac operator equals the optimal rational approximation to the overlap operator used in the simulation.","marker":"[8]"},{"why":"Supplies the ensemble A of 535 configurations and the lattice parameters used throughout the analysis.","marker":"[9]"},{"why":"Provides the domain-wall fermion framework that underlies exact chiral symmetry on the lattice.","marker":"[5]"},{"why":"Documents the eigenmode-projection procedures used to obtain overlap zero modes and low-lying modes.","marker":"[11]"},{"why":"Supplies the adaptive thick-restart eigensolver used for projecting the low-lying modes.","marker":"[10]"},{"why":"Establishes, together with the following reference, the plaquette criterion for topological sectors used to test flowed configurations.","marker":"[13]"},{"why":"Provides the mathematical result used with the preceding reference to formulate the topological-sector condition.","marker":"[14]"}],"fun_headline_variants":["Asymptotic clover charge reproduces QCD topological susceptibility","Wilson flow clover charge gives same QCD susceptibility as overlap index","Rounded clover charge from flow matches overlap index for QCD topology","Topological susceptibility from clover flow charge equals overlap-index value","Exact chiral QCD: flowed clover charge substitutes overlap index for topology"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that the integer value the rounded clover charge settles on at large Wilson-flow time is the true topological charge of the underlying gauge field, even though 31.2% of the configurations change their integer charge during the flow relative to the overlap index at $t=0$.","fun_headline_variants_meta":{"raw":{"variants":["Asymptotic clover charge reproduces QCD topological susceptibility","Wilson flow clover charge gives same QCD susceptibility as overlap index","Rounded clover charge from flow matches overlap index for QCD topology","Topological susceptibility from clover flow charge equals overlap-index value","Exact chiral QCD: flowed clover charge substitutes overlap index for topology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000767,"raw_usage":{"total_tokens":3521,"prompt_tokens":1187,"completion_tokens":2334,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":803,"completion_tokens_details":{"reasoning_tokens":2252}},"tokens_in":803,"tokens_out":2334,"duration_ms":16152,"temperature":1.0,"reasoning_tokens":2252,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:06:46.093654+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"On this same ensemble, compute the fourth cumulant $c_4$ of the topological charge distribution from the overlap index and from $Q_{tc}$. If they disagree beyond statistical error, the claim that $Q_{tc}$ captures the full topological fluctuations of the vacuum is false, because $c_4$ is part of the distribution the paper claims to reproduce.","supporting_citations":[{"cited_title":"Yamazaki, Z","cited_arxiv_id":null,"evidence_quote":"Supplies the adaptive thick-restart eigensolver used for projecting the low-lying modes."},{"cited_title":"Luscher, Commun","cited_arxiv_id":null,"evidence_quote":"Establishes, together with the following reference, the plaquette criterion for topological sectors used to test flowed configurations."},{"cited_title":"Phillips and D","cited_arxiv_id":null,"evidence_quote":"Provides the mathematical result used with the preceding reference to formulate the topological-sector condition."}],"review_version":1}