{"id":"ab37e778-39c2-43f5-96d3-72d483895c65","arxiv_id":"2505.10295","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"At T≈254 MeV, lattice QCD gives -[HE(ω2)-HE(ω1)]/T^2 = 0.193(74), on the low side of the AMY weak-coupling prediction interval [0.25,0.30].","lead":"Lattice QCD was used to compute two moments of the quark-gluon plasma's photon spectrum at about 254 MeV without solving an inverse problem. The hard-photon moment difference is lower than, but compatible with, the leading-order weak-coupling prediction, offering a new non-perturbative input to the direct photon puzzle.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed compatibility with the AMY photon spectrum appears to use the isospin-current moment without applying the paper's own conversion σ_em≃(2/3)σ_iso; the converted value is not compatible with AMY.","rationale":"The reader identified the tail-extension ansatz as the weakest assumption, but that is a systematic-uncertainty issue that is at least explicitly assigned an error (the 0.044 tail systematic). The isospin-to-EM conversion is a factor-of-3/2 discrepancy that directly invalidates the abstract's central claim of compatibility with the weak-coupling photon spectrum. The factor is stated in the paper's own footnote but not applied to the quoted numbers, making this a concrete, checkable arithmetic issue rather than a modeling uncertainty. If the factor is indeed missing, the paper's conclusion changes from 'compatible with AMY' to a significant discrepancy in the opposite direction, which is a major alteration of the central physical message. The numerical lattice result may still be valid as a measurement of the isospin-current moment, but the interpretation as a constraint on the photon spectrum would need substantial revision, hence a conditional verdict rather than outright rejection.","tokens_in":10749,"tokens_out":19599,"duration_ms":151685,"concrete_test":"Recompute the AMY moment using the isospin current normalization (replace Σ_f q_f^2 = 2/3 by 1 in the AMY spectral function) and compare with Eq. (18); equivalently, multiply Eq. (18) by 2/3 and test the difference from the quoted [0.25,0.30]. If the converted value 0.129(49) lies more than 2σ from the AMY interval, the compatibility claim in the abstract fails.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The lattice calculation in Eq. (4) and the final result Eq. (18) are evaluated with the isospin current jμ=(uγμu−dγμd)/√2, which has charge-squared sum 1. The AMY spectral function [17] is for the electromagnetic current, whose charge-squared sum is 2/3 for three massless flavors (u,d,s). The paper's own footnote states σ_em ≃ (4/9+1/9+1/9)σ = (2/3)σ_iso. Therefore the physical photon-spectrum moment corresponding to Eq. (18) is (2/3)×0.193 = 0.129(49), which is more than 2σ below the AMY interval [0.25,0.30]. Calling this 'lower than, but compatible' is not supported. Conversely, if the comparison were meant to be for the isospin current, the AMY interval would need to be multiplied by 3/2, giving [0.375,0.45], and 0.193(74) would again be about 2.5σ below. Either way, the central comparison in the abstract and conclusion does not hold. The 2.6σ evidence for a nonzero moment survives the rescaling (0.129/0.049 ≈ 2.6), but the interpretation as a direct probe of the photon spectrum and the relevance to the direct photon puzzle are materially affected.","agreement_with_reader":"disagree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper computes two moments of the thermal photon spectral function, specifically the difference -[H_E(ω2)-H_E(ω1)]/T², at T≈254 MeV using lattice QCD with Nf=2 O(a)-improved Wilson fermions. The calculation is performed by evaluating the screening correlators in the second Matsubara sector and extending them beyond the reference distance x1=β with a finite-exponential ansatz, thereby avoiding a numerically ill-posed inverse problem. The central result is -[H_E(ω2)-H_E(ω1)]/T² = 0.193(74), which the authors describe as lower than, but compatible with, the leading-order AMY weak-coupling prediction of 0.25–0.30, and as 2.6σ evidence for a non-zero photon emissivity. The paper also presents a detailed error budget separating statistical, continuum-extrapolation, and tail-systematic uncertainties, with model averaging over twelve continuum fits and with/without-prior tail analyses.","tokens_in":11062,"tokens_out":7413,"duration_ms":71891,"significance":"If the computation is correct, this is a significant methodological advance: it provides a first-principles lattice constraint on hard-photon emission from the quark-gluon plasma without solving an inverse problem, with high statistics and a transparent systematic error budget. The direct comparison with the leading-order weak-coupling spectrum is an important benchmark for the direct-photon puzzle in heavy-ion phenomenology. The paper deserves credit for the improved statistical precision, the model-averaged continuum extrapolation, and the explicit treatment of the tail extension. However, the headline comparison with the AMY spectrum is affected by a normalization inconsistency between the lattice isospin current and the electromagnetic current, as detailed below; this changes the conclusion about compatibility but does not invalidate the numerical calculation or the evidence for a non-zero emissivity.","major_comments":[{"comment":"The comparison with the AMY interval is made with inconsistent current normalizations. The lattice quantity in Eq. (18) is computed with the isospin current jμ=(ūγμu−d̄γμd)/√2, whose charge-squared sum is one, whereas the AMY spectral function entering Eq. (1) is for the electromagnetic current with charge-squared sum 2/3 for three flavors. Footnote 2 itself states σ_em≃(2/3)σ_iso. Thus the physical photon-spectrum moment corresponding to Eq. (18) is (2/3)×0.193(74)=0.129(49), which is 2.5σ–3.5σ below the quoted AMY interval [0.25,0.30], not compatible with it. Equivalently, comparing the isospin moment with 3/2 times the AMY interval gives [0.375,0.45], again about 2.5σ above the lattice value. The abstract and conclusion must be revised: the lattice result is significantly lower than the leading-order weak-coupling prediction once the normalization is applied. The 2.6σ evidence for a non-zero moment is unaffected because it is a test against zero.","section":"Results (Eq. (18)) and footnote 2"},{"comment":"The tail systematic is taken as the full difference between the with-prior and without-prior results, but both analyses employ the same finite-sum ansatz Eq. (9) with at most three states. This difference therefore reflects sensitivity to priors and fit ranges, not the error incurred if the true large-x1 correlator contains additional states or a continuum contribution. Since the tail extension beyond x1=β contributes directly to H_E(ω2), the authors should either bound the omitted higher-state contribution using the fitted amplitudes and masses, or enlarge the tail systematic to cover the model uncertainty. The one-, two-, and three-state comparison in Fig. 6 checks the ground-state mass, not the tail integral, so it does not by itself close this gap.","section":"Results (Eq. (9)) and End Matter 'Correlator fits'"}],"minor_comments":[{"comment":"The interpolation parameters xw and d are not fully specified in the text or the figure caption; the text mentions d≈0.15 fm but not the value of xw. Please state explicitly that the transition is centered at x1=β or give the numerical value.","section":"Eq. (19) and Fig. 4"},{"comment":"The phrase 'two moments of that spectrum' is imprecise because the quoted result in Eq. (18) is the difference of two moments, not the moments themselves; please rephrase for precision.","section":"Abstract"},{"comment":"Reference [17] is missing the publication year; please add it.","section":"References"},{"comment":"The statement that the calculation is 'free of systematic uncertainties associated with an inverse problem' is potentially too strong, since the tail extension in Eq. (9) is a modeling assumption; consider saying 'free of an inverse-problem reconstruction' or similar.","section":"Results, final paragraph"},{"comment":"The columns for the number of inversions would be clearer with explicit subscripts ω1 and ω2 in the header rather than the current formatting.","section":"End Matter, Table I"}],"recommendation":"major_revision","confidential_remarks":"The numerical calculation is careful and the methodological advance is real. The normalization inconsistency in the AMY comparison is fixable but changes the headline claim: after applying the paper's own factor 2/3, the lattice result is not compatible with the leading-order weak-coupling prediction. The nonzero-emissivity evidence and the direct-photon-puzzle relevance can still be presented, but the abstract and conclusion need substantial rewording. I recommend major revision rather than rejection because the lattice computation itself appears sound and the issues are correctable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The lattice work behind this letter is real and demanding: the n=2 moment of the quark-gluon plasma spectral function at T≈254 MeV is computed directly from first principles, no inverse problem, with a statistical gain up to a factor 6.5 in the integrand. The tail-systematics study, the 12-model continuum extrapolation, and the with/without-prior bracket are all careful. The central number for the isospin current, -[HE(ω2)-HE(ω1)]/T^2 = 0.193(74), is a new result and worth a serious referee.\n\nThe problem is the comparison to AMY. The lattice computation uses the isospin current, with charge-squared sum 1. The AMY spectral function is for the electromagnetic current, charge-squared sum 2/3 for u,d,s. The paper's own footnote says σ_em ≃ (2/3)σ_iso. Consistent conversion gives the lattice EM moment as 0.193×(2/3)=0.129(49). The quoted AMY interval [0.25,0.30] is EM-normalized, so the difference is 2.5–3.5σ, not compatible. If you instead keep both in isospin normalization, the AMY interval becomes [0.375,0.45], and 0.193(74) is still low by roughly 2.5σ. Either way, the abstract's 'lower than, but compatible' is not supported by the numbers.\n\nThis is not a cosmetic issue. The paper's scientific message—that lattice QCD is consistent with leading-order weak-coupling emissivity—is the opposite of what the numbers show once the normalization is applied. The 2.6σ evidence for a nonzero moment survives the rescaling, but the conclusion should say the lattice moment is significantly below AMY, which strengthens the direct-photon puzzle instead of resolving it. The fix is mechanical: apply the factor 2/3 consistently and rewrite the comparison.\n\nEverything else checks out. The finite-state ansatz for the tail is a reasonable assumption, honestly treated as a systematic; the no-prior/prior difference covers it rather than hiding it. The EQCD agreement for the non-static screening mass adds confidence. Data and code are not public, which is typical for this kind of lattice letter and limits independent audit but is not a red flag.\n\nRecommendation: send it to referees. The computational result is valuable, and the normalization error is correctable. But flag the AMY conversion to the referees, because the paper in its current form overstates the agreement with weak coupling.","headline":"A careful lattice determination of the n=2 photon-spectral moment, but the AMY comparison ignores the isospin-to-EM normalization factor of 2/3, so the claimed compatibility does not hold.","tokens_in":11611,"tokens_out":5405,"would_cite":true,"duration_ms":50065,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81V05","81T25","81T80"],"pacs":[],"model":"deepseek-v4-flash","headline":"Lattice QCD measures the quark-gluon plasma's hard-photon emissivity without an inverse problem and finds 0.193(74) for the two-moment difference.","keywords":["lattice QCD","thermal photon emissivity","quark-gluon plasma","spectral function moments","direct photon puzzle","screening correlators","inverse problem","Wilson fermions"],"falsifier":"Measure the $n=2$ transverse screening correlators at separations far beyond $x_1=\\beta$ with sufficient precision to see the tail directly; if the ground-state mass gap disagrees with Eqs.~(10)--(13) by more than the quoted errors, or a negative-amplitude state is required, then $-[H_E(\\omega_2)-H_E(\\omega_1)]/T^2$ will move outside $0.193(74)$.","tokens_in":10565,"feed_emoji":"","tokens_out":9296,"duration_ms":82188,"temperature":0.7,"pith_summary":"This paper aims to establish that the hard-photon part of the quark-gluon plasma's thermal photon spectrum can be computed directly from lattice QCD, without solving an ill-posed inverse problem. At $T\\approx 254\\,\\mathrm{MeV}$ it evaluates the second Matsubara moment of the photon spectral function and combines it with the previously determined first moment. The difference, $-[H_E(\\omega_2)-H_E(\\omega_1)]/T^2$, is $0.193(74)$; it suppresses soft photons, isolates energies $\\omega\\gtrsim\\pi T\\approx 1\\,\\mathrm{GeV}$, and is $2.6\\sigma$ away from zero. It lies below the leading-order weak-coupling interval $[0.25,0.30]$, and if correct it gives heavy-ion phenomenology a non-perturbative anchor for thermal photon emission, where the direct photon puzzle currently forces models to stretch.","feed_headline":"Hard-photon glow of quark-gluon plasma measured on the lattice","feed_subtitle":"A lattice-QCD moment difference gives 0.193(74) at 254 MeV, below but compatible with weak-coupling predictions.","key_machinery":"The central object is the difference of two spectral-function moments, $H_E(\\omega_2)-H_E(\\omega_1)$, where $H_E(\\omega_n)$ is the fixed-lightlike-virtuality correlation function of the vector current defined by Eq.~(2). On the lattice these moments are expressed as sums over transverse static and non-static screening correlators, with a kernel $\\Omega_n$ chosen (via a one-parameter family and lattice perturbation theory) to suppress $O(a^2)$ artifacts. The $n=2$ integrand's exponentially deteriorating signal-to-noise ratio is handled with stochastic wall sources and a truncated solver, reducing integrand errors by up to a factor of about 6.5, and by fitting the long-distance tail as a sum of exponentials with positive prefactors (Eq.~(9)) in order to extend the integrand beyond $x_1=\\beta$. Akaike-weighted model averaging over continuum extrapolations and kernel choices sets the central value and error, and subtracting the precisely known first moment removes the soft-photon contribution.","core_discovery":"On the paper's own terms, the central result is Eq.~(18): at $T\\approx 254\\,\\mathrm{MeV}$ the difference of the $n=2$ and $n=1$ moments of the photon spectral function is $-[H_E(\\omega_2)-H_E(\\omega_1)]/T^2 = 0.193(74)$, computed directly from the lattice with two flavors of $O(a)$-improved Wilson fermions. The paper argues this is the first such moment computation free of the systematic uncertainties of an inverse-problem spectral reconstruction, and that the positive value constitutes $2.6\\sigma$ evidence that the quark-gluon plasma emits hard photons at this temperature. The quoted value is lower than, but compatible with, the leading-order weak-coupling kinetic-theory result, which lies in $[0.25,0.30]$ for $\\alpha_s\\in[0.25,0.31]$.","pith_inferences":["If the same two moments were computed at a second temperature, for instance near the crossover, the change in the difference would test whether hard-photon emission strengthens as $T$ approaches $T_c$, as the first moment alone already hints.","A third moment, obtained by the same no-inversion route, would begin to constrain the shape of the hard-photon spectrum rather than only its normalization.","The gap between the with-prior and without-prior tail treatments suggests that a direct, high-precision measurement of the $n=2$ screening correlators beyond $x_1=\\beta$ is the fastest way to shrink the dominant systematic uncertainty."],"forward_implications":["Hard-photon emissivity of quark-gluon plasma near $T\\approx 1.2T_c$ is now pinned from first principles, so thermal-photon predictions in heavy-ion models no longer need to rely solely on weak-coupling input.","Because the result sits below the leading-order weak-coupling band and the next-to-leading-order correction is positive, improving the perturbative comparison at this temperature widens rather than closes the gap.","The moment-difference method gives a route to other transport quantities such as electric conductivity with controlled systematic errors and no inverse problem.","The non-zero value provides a concrete target for hydrodynamic calculations of direct-photon yield and azimuthal anisotropy at RHIC and the LHC."],"supporting_citations":[{"why":"Sets the temperature scale, identifying $T\\approx 254\\,\\mathrm{MeV}$ as roughly $1.2T_c$ for the ensembles.","marker":"[16]"},{"why":"Supplies the leading-order weak-coupling photon spectrum whose moments are the comparison target.","marker":"[17]"},{"why":"Determines the first moment $H_E(\\omega_1)$ used in the subtraction.","marker":"[23]"},{"why":"Establish the stochastic wall-source techniques used to reach the high statistics on the screening correlators.","marker":"[27-29]"},{"why":"Supplies the stochastic momentum source method used in the correlator computation.","marker":"[30]"},{"why":"Provides the truncated solver approach that makes the many low-precision solves efficient.","marker":"[31]"},{"why":"Gives the Akaike information criterion used for model weights in the averaging.","marker":"[33]"},{"why":"Provide the model-averaging method used to combine continuum-extrapolation models.","marker":"[34,35]"},{"why":"Provides the EQCD prediction for the non-static screening mass in the $\\omega_2$ sector, used to validate the tail fit.","marker":"[40]"},{"why":"Computes the next-to-leading-order correction to the weak-coupling emissivity, stated to be positive in the interpretation.","marker":"[41]"}],"fun_headline_variants":["First lattice-QCD photon moments without inverse problem","Lattice QCD: QGP photon moment difference 0.193(74) at 254 MeV","Quark-gluon plasma photon spectrum moments from lattice QCD","Lattice QCD photon moments match weak-coupling prediction for QGP","Hard-photon emission from QGP quantified by lattice QCD moments"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All quoted numbers rest on the assumption that, beyond $x_1=\\beta$, the static and non-static screening correlators are exactly described by a finite sum of exponentials with positive prefactors (Eq.~(9)), so replacing the measured tail by the fitted tail does not bias the moment.","fun_headline_variants_meta":{"raw":{"variants":["First lattice-QCD photon moments without inverse problem","Lattice QCD: QGP photon moment difference 0.193(74) at 254 MeV","Quark-gluon plasma photon spectrum moments from lattice QCD","Lattice QCD photon moments match weak-coupling prediction for QGP","Hard-photon emission from QGP quantified by lattice QCD moments"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000276,"raw_usage":{"total_tokens":1616,"prompt_tokens":887,"completion_tokens":729,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":503,"completion_tokens_details":{"reasoning_tokens":633}},"tokens_in":503,"tokens_out":729,"duration_ms":7500,"temperature":1.0,"reasoning_tokens":633,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T21:12:59.465512+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the $n=2$ transverse screening correlators at separations far beyond $x_1=\\beta$ with sufficient precision to see the tail directly; if the ground-state mass gap disagrees with Eqs.~(10)--(13) by more than the quoted errors, or a negative-amplitude state is required, then $-[H_E(\\omega_2)-H_E(\\omega_1)]/T^2$ will move outside $0.193(74)$.","supporting_citations":[{"cited_title":"Chatterjee and P","cited_arxiv_id":null,"evidence_quote":"Sets the temperature scale, identifying $T\\approx 254\\,\\mathrm{MeV}$ as roughly $1.2T_c$ for the ensembles."},{"cited_title":"ratio method","cited_arxiv_id":null,"evidence_quote":"Determines the first moment $H_E(\\omega_1)$ used in the subtraction."}],"review_version":1}