{"id":"13b21931-9900-484e-ba6b-71a28b884d8d","arxiv_id":"2411.14127","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":3.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Non-perturbative lattice QCD shows hadronic screening masses deviate from NLO perturbation theory from 1 to 160 GeV, with the baryonic NLO correction computed for the first time.","lead":"Physicists used supercomputers to simulate the quark-gluon plasma at temperatures up to 160 GeV, far hotter than any previous lattice calculation. The results show that simple perturbative formulas fail to describe how hadrons screen in the plasma across the entire temperature range, a strong hint that fully non-perturbative methods are needed even at electroweak-scale temperatures.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The NLO-failure claim is sensitive to the choice of renormalization scale used to define ghat^2; scale variation can shift the NLO curve by O(g^4), comparable to the fitted quartic terms, so a scale-robustness test is needed before interpreting the data as evidence against perturbative QCD at…","rationale":"I read this as a proceedings summarizing a credible multi-year program. The continuum extrapolations, the use of shifted boundary conditions, and the step-scaling renormalization described in Section 3 are supported by the cited published papers, and I do not question the numerical data themselves. The central interpretive claim, however, is that the NLO perturbative expressions in eqs. (13) and (16) fail over the entire temperature range, and that this indicates limited applicability of perturbation theory even at the electroweak scale. That claim depends on the convention used to define ghat^2. The NLO coefficient is only meaningful once a renormalization scale is chosen; at the finite couplings probed here, a change of scale from pi T to 4 pi T changes the NLO prediction by O(g^4), which is the same order as the fitted quartic terms. The paper's own admission in Section 5.2 that alternative parameterizations can make b2 disagree with the NLO value strengthens this concern. My proposed test, varying the scale and if needed the loop order of the coupling, would settle whether the observed curvature is physical evidence against NLO perturbation theory or an artifact of the chosen comparison. The reader's primary weakest assumption, the zero-topology restriction, is less load-bearing for the central claim because in the massless-quark limit topological sectors are heavily suppressed in the continuum; the scale-convention issue is the more decisive check. Since this concern does not invalidate the measured masses but does affect the interpretation of the fitted higher-order terms, the appropriate verdict remains conditional, matching the reader's assessment.","tokens_in":14117,"tokens_out":7874,"duration_ms":82935,"concrete_test":"Using the published continuum-limit meson and baryon screening masses, repeat the fits of Sections 5.1 and 5.2 with ghat^2 defined at mu = pi T and mu = 4 pi T, and also with three-loop running, again fixing p0, p2 and b0, b2 to the NLO values in eqs. (13) and (16). If the NLO curves are more than 3 sigma away from the data at every scale choice, the NLO-failure claim survives; if any scale choice brings the NLO curves within errors, the claim is a scale-convention artifact and should be weakened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim rests on comparing continuum-extrapolated screening masses to the NLO expressions in eqs. (13) and (16), evaluated with the two-loop coupling ghat^2 defined at mu = 2 pi T (eq. (17)). Because the data are at finite g^2 over the range, and because the NLO curve is specified only up to O(g^4), changing the renormalization scale to pi T or 4 pi T shifts the NLO prediction by O(g^4). These shifts are of the same size as the fitted quartic coefficients, e.g. p4 = -0.0161(17) and b4 = -0.021(3). The paper does not vary the scale or demonstrate that no reasonable scale choice brings the NLO curves into agreement; instead it fixes p2 and b2 to the perturbative values and fits g^3 and g^4 terms. The paper itself notes in Section 5.2 that alternative parameterizations are possible and make b2 disagree with eq. (16), which underscores that the conclusion is tied to this convention. Thus the statement that NLO perturbation theory is insufficient is not yet robust against the dominant theoretical systematic of the comparison, even though the underlying lattice data and continuum extrapolations are credible. The zero-topology restriction in Section 3.4 is a lesser concern here, since massless quarks suppress nonzero topology in the continuum, although lattice-artifact contamination at T ~ 1 GeV should still be checked.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper reports a strategy for non-perturbative lattice QCD at very high temperatures (T ≈ 1–160 GeV) by combining Schrödinger-functional step-scaling renormalization with shifted boundary conditions. It presents the hadronic screening spectrum (mesons and baryons) at twelve temperatures with continuum extrapolations, and compares the results with the NLO perturbative expressions in eqs. (13) and (16), evaluated with the two-loop coupling ĝ²(2πT) defined in eq. (17). The paper claims that the NLO terms alone cannot describe the data and that higher-order (ĝ³ and ĝ⁴) contributions are required over the entire temperature range. It also reports the first quantitative NLO calculation of the baryonic screening mass in the three-dimensional effective theory, cross-checked by two independent numerical methods in appendix A.","tokens_in":14412,"tokens_out":10227,"duration_ms":99843,"significance":"The computational strategy is a substantial methodological advance: it avoids the scale-separation problem of standard thermal lattice simulations and yields permille-level continuum screening masses across more than two orders of magnitude in temperature. If the interpretive claim is robust, these results provide the first non-perturbative evidence on the slow convergence of the weak-coupling expansion for screening masses at electroweak-scale temperatures, and the NLO baryonic result in eq. (16) is a useful new perturbative input. The paper is honest about the parameterization dependence of its fits (Section 5.2), but for that same reason the central claim needs additional robustness tests before publication.","major_comments":[{"comment":"The central claim that the NLO expressions are insufficient is calibrated to ĝ² evaluated at μ = 2πT. Since eq. (17) is the two-loop running coupling, changing the renormalization scale to μ = πT or 4πT shifts ĝ² by O(g⁴); translating this shift to the screening-mass ratio changes the effective quartic coefficient by roughly 0.003–0.005 for an order-of-magnitude change in μ. This is a substantial fraction of the fitted coefficients p₄ = −0.0161(17) and b₄ = −0.021(3), and several times their statistical errors. The paper fixes p₂ and b₂ to the perturbative values and never varies the scale, so the statement that higher-order terms are required is not yet established independently of this convention. Please refit with μ = πT and 4πT (and, if possible, with p₂ and b₂ left free) and report whether the need for g³/g⁴ terms persists; if different reasonable scale choices change the significance or sign of the quartic terms, the conclusion should be weakened accordingly.","section":"Secs. 5.1–5.2, eqs. (13), (16), (17)–(21)"},{"comment":"The fits impose the NLO coefficients p₂ and b₂ rather than determining them from the data. The paper itself notes that other parameterizations of the baryonic data make b₂ disagree with eq. (16), which means the reported b₃, b₄ (and similarly p₃, p₄) are conditional on the assumed validity of the NLO coefficient—precisely the object under test. A more robust analysis would fit p₂ and b₂ freely and compare models with and without g³/g⁴ terms using an information criterion or an F-test; at minimum, the free-fit results and the resulting χ² values should be reported so the reader can see whether the NLO coefficient is actually compatible with the data when not enforced.","section":"Sec. 5.2, eq. (21); Sec. 5.1.1, eq. (18)"},{"comment":"The restriction to the zero-topological-charge sector is justified by a dilute-instanton-gas estimate, but no numerical check is provided at the lowest simulated temperature T ≈ 1 GeV. Since the simulated theory has massless quarks, the continuum topological susceptibility is expected to vanish, but at finite lattice spacing with Wilson fermions the suppression of nonzero topology can be weaker, and a residual O(a) or finite-volume contamination would not be removed by the continuum extrapolation. This is an unverified assumption in the data pipeline for the lowest-temperature points, which help determine p₄ and b₄. The authors should either provide a direct check (for example, measuring the topological charge distribution at the smallest T, or comparing screening masses with and without the zero-topology restriction) or quantify the expected contamination at T ≈ 1 GeV.","section":"Sec. 3.4"}],"minor_comments":[{"comment":"The definition I₁(x₃,L) ≡ (1 − lim_{x₃→∞}) C_O(x₃) is not mathematically well-formed as written: taking the limit before subtracting from C_O(x₃) would give a constant rather than the x₃-dependent residue described in the text. Presumably the intended definition is a normalized residue such as I₁ = 1 − C_O(x₃)/lim_{x₃→∞} C_O(x₃).","section":"Eq. (9)"},{"comment":"The reported covariance ratios contain index mismatches: cov(p₃,s₄)/[σ(p₃)σ(p₄)] should presumably use σ(s₄) in the denominator, and cov(p₄,s₄)/[σ(p₄)σ(p₄)] should presumably be cov(p₄,s₄)/[σ(p₄)σ(s₄)].","section":"Sec. 5.1.2, text after eq. (20)"},{"comment":"Typo: “necessery” should be “necessary”.","section":"Sec. 5.1.1"},{"comment":"The abstract says “the known leading behaviour in the coupling constant,” but eqs. (13) and (16) give the next-to-leading (O(g²)) correction after the free-theory term; the wording should be “first non-trivial interacting contribution” or “next-to-leading” for consistency with the body.","section":"Abstract and Sec. 6"},{"comment":"The estimate is phrased for “three light degenerate flavours of mass m,” while the simulations are described elsewhere as having exactly massless quarks; the instanton-gas argument should be stated for the massless case actually simulated, or the discrepancy should be explained.","section":"Sec. 3.4"}],"recommendation":"major_revision","confidential_remarks":"This is a proceedings contribution that largely summarizes results published in Refs. [7] and [8]; the editor may wish to consider whether the scale-robustness analysis and the free-coefficient fits can be added within the proceedings format. The underlying lattice data and the NLO baryonic computation appear credible, but the paper's advertised conclusion about the failure of NLO perturbation theory is currently too dependent on the choice of renormalization scale and on the imposed NLO coefficients to be accepted as stated."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Best to read this as a status report on work you should already know from Refs. [7] and [8]. The genuinely new content is the synthesis and the polynomial parameterizations, not the underlying numbers. That said, what the group has done is substantial: continuum-extrapolated screening masses at few-permille accuracy from 1 to 160 GeV with N_f = 3, plus the first quantitative NLO baryonic screening mass via two independent numerical solutions of the effective-field-theory Schrödinger equation. The strategy—step scaling plus shifted boundary conditions—is clever, and this paper explains it cleanly. Credit where due: the methods sections are clear, the fits have good chi^2, and the quoted statistical errors are believable.\n\nThe soft spot is the interpretive step. The claim that NLO perturbation theory is insufficient rests on comparing data to eqs. (13) and (16) evaluated at mu = 2 pi T via the two-loop coupling of eq. (17). The NLO curve is defined only up to O(g^4), and changing the scale to pi T or 4 pi T shifts the NLO prediction by O(g^4)—the same size as the fitted p4 and b4 coefficients. The paper fixes p2 and b2 to the perturbative values and then fits g^3 and g^4 terms. That is a legitimate parameterization, but the conclusion that NLO fails is tied to that convention. The paper itself admits alternative parameterizations are possible and would make b2 disagree with eq. (16). So the claim would be strengthened by a scale-variation study: if no reasonable scale choice brings NLO into agreement with the data, then the conclusion is robust. As written, the headline overinterprets a convention-dependent comparison.\n\nThe zero-topology restriction (Sec. 3.4) is less worrying, since massless quarks suppress nonzero topology in the continuum and the dilute instanton gas estimate supports the approximation. Still, lattice-artifact contamination at T ~ 1 GeV could be checked. Minor.\n\nThis is a proceedings paper. As such, it is solid and worth reading. Do not expect new physics results here; the primary results are in Refs. [7] and [8]. For peer review: if this venue reviews proceedings, accept with a request to add a scale-robustness discussion and to soften the “NLO fails” wording. If it is a talk summary, the current form is acceptable.","headline":"A solid proceedings summarizing a first-class lattice program, but the claim that NLO perturbation theory fails is tied to a fixed renormalization-scale convention and needs a scale-robustness check before it stands as stated.","tokens_in":15005,"tokens_out":1936,"would_cite":false,"duration_ms":17575,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T25","81V05"],"pacs":["11.15.Ha","12.38.Gc"],"model":"deepseek-v4-flash","headline":"A first non-perturbative lattice computation of hadronic screening masses in three-flavour QCD from 1 to 160 GeV shows the known next-to-leading-order perturbative formulas are insufficient across the entire range.","keywords":["lattice QCD","thermal QCD","screening masses","shifted boundary conditions","step scaling","electroweak scale","perturbative expansion","dimensional reduction"],"falsifier":"Compute the topological-charge distribution on the largest spatial-volume lattices at $T\\simeq1$ GeV, where the paper's boxes have $LT$ between 20 and 50, and compare screening masses extracted with and without the $Q=0$ restriction; if configurations with $|Q|>0$ occur with probability near or above $10^{-3}$, the continuum screening masses would shift by more than their quoted permille accuracy.","tokens_in":13854,"feed_emoji":"♨️","tokens_out":15546,"duration_ms":134143,"temperature":0.7,"pith_summary":"This paper reports the first non-perturbative lattice computation of hadronic screening masses in three-flavour QCD across a temperature range that reaches the electroweak scale, from roughly 1 GeV to about 160 GeV. The authors argue that the known next-to-leading-order perturbative formulas, $m/(2\\pi T)=1+0.0327\\,g^2$ for mesons and $m_N/(3\\pi T)=1+0.046\\,g^2$ for baryons, are not enough to describe their continuum-extrapolated data anywhere in this range. Instead, cubic and quartic terms in the running coupling are needed, so perturbation theory remains incomplete even at electroweak-scale temperatures. This matters because it establishes both a practical first-principles route to high-temperature QCD and a precise target for improved finite-temperature perturbation theory.","feed_headline":"QCD screening masses up to 160 GeV need more than the known formula","feed_subtitle":"First non-perturbative calculation from 1 to 160 GeV shows higher-order corrections persist across the full range.","key_machinery":"The argument is carried by a computational strategy for simulating thermal QCD without placing the pion mass and the temperature on the same lattice. Step-scaling techniques define a non-perturbative running coupling in a finite volume, fixing the lines of constant physics at a lattice spacing set by the temperature; shifted boundary conditions remove the need for a zero-temperature subtraction and, via effective-field-theory arguments, keep finite-volume effects exponentially small. The screening masses are extracted from the exponential falloff of spatial two-point correlation functions in the zero-topological-charge sector, and the temperature dependence is analysed with the two-loop running coupling $\\hat g^2(T)$ at scale $\\mu=2\\pi T$ with $\\Lambda_{\\overline{\\rm MS}}=341$ MeV. On the perturbative side, the baryonic next-to-leading-order prediction is obtained by solving a $(2+1)$-dimensional quantum-mechanical eigenvalue problem for the three-quark system in the dimensionally reduced effective theory (electrostatic QCD plus non-relativistic QCD), with the static potentials $V_\\pm(r)$ as input.","core_discovery":"The central discovery is that, in $N_f=3$ QCD, hadronic screening masses computed non-perturbatively with continuum-limit extrapolations and few-permille accuracy from $T=1$ GeV to $T\\simeq160$ GeV do not follow their next-to-leading-order perturbative predictions. The pseudoscalar mass is described over the whole range by $m_P/(2\\pi T)=p_0+p_2\\hat g^2+p_3\\hat g^3+p_4\\hat g^4$, with $p_0=1$ and $p_2=0.0327$ fixed to perturbation theory and fitted $p_3=0.0038(22)$, $p_4=-0.0161(17)$; the vector channel adds a spin-dependent term $s_4\\hat g^4$ with $s_4=0.00704(14)$, so the vector--pseudoscalar splitting is nonzero up to 160 GeV. Positive and negative parity nucleon screening masses are degenerate throughout, as expected from chiral symmetry restoration. The nucleon screening mass is parameterized as $m_{N+}/(3\\pi T)=b_0+b_2\\hat g^2+b_3\\hat g^3+b_4\\hat g^4$, with $b_0=1$ and $b_2=0.046$ fixed and $b_3=0.026(4)$, $b_4=-0.021(3)$. On this basis the paper concludes that higher-order terms in the running coupling remain relevant across the whole range and that next-to-leading-order perturbation theory is not sufficient. The paper also notes that other parameterizations of the baryonic data are possible and would move the fitted NLO coefficient away from the perturbative value, so the quoted cubic and quartic coefficients are tied to the chosen polynomial ansatz.","pith_inferences":["If the same pattern carries over to other quantities, electroweak-scale QCD inputs used in early-universe cosmology, such as the equation of state and transport coefficients, may also deviate from their perturbative estimates.","The fitted coefficients $p_3$, $p_4$, $b_3$, $b_4$ are concrete targets for a future full next-to-next-to-leading-order calculation in the three-dimensional effective theory; agreement would validate dimensional reduction at these scales, while disagreement would point to contributions outside that theory.","The step-scaling plus shifted-boundary combination is not obviously specific to QCD and could be adapted to other asymptotically free gauge theories, or to QCD with different flavour numbers, where electroweak-scale thermodynamics may be needed.","The paper's zero-topology restriction could be checked by measuring the topological susceptibility on the same lattices at $T\\simeq1$ GeV, since the safety argument is semi-classical rather than a direct lattice measurement at this exact setup."],"forward_implications":["If the paper is right, next-to-leading-order perturbation theory cannot be used to predict screening masses at any temperature from 1 to 160 GeV; the quartic term is needed even at the top of the range.","The observed $\\hat g^4$ scaling of the vector--pseudoscalar splitting, with a nonzero coefficient at the highest simulated temperature, means spin-dependent effects missed by the next-to-leading-order computation are visible across the whole range.","For the nucleon, the fitted polynomial gives $m_{N+}/(3\\pi T)=1+0.046\\,\\hat g^2+0.026\\,\\hat g^3-0.021\\,\\hat g^4$, with positive and negative parity masses degenerate over the entire interval.","The same combination of step scaling and shifted boundary conditions opens the way to non-perturbative results for other thermal observables, such as the equation of state, up to the electroweak scale at moderate computational cost.","The data reinforce the paper's conclusion that perturbative thermal QCD has poor convergence at finite temperature even at 160 GeV, so a fully non-perturbative treatment is needed."],"supporting_citations":[{"why":"It supplies the step-scaling and shifted-boundary strategy and the mesonic screening-mass data and fits that are central to this paper.","marker":"[7]"},{"why":"It supplies the baryonic screening-mass lattice results and their continuum extrapolations, which are a main new result reported here.","marker":"[8]"},{"why":"It provides the next-to-leading-order mesonic screening-mass expression, eq. (13), against which the non-perturbative data are compared.","marker":"[9]"},{"why":"It provides the next-to-leading-order baryonic screening-mass calculation in the effective theory, eq. (16), obtained from the three-body quantum-mechanical problem.","marker":"[10]"},{"why":"It introduced shifted boundary conditions for the equation of state in SU(3) Yang-Mills theory, establishing the simulation setup used for the thermal ensembles.","marker":"[6]"},{"why":"It derives the equivalence of shifted boundary conditions with a longer compact direction and the exponential finite-volume suppression used to justify the large spatial volumes.","marker":"[27]"},{"why":"It determines the three-flavour QCD scale $\\Lambda_{\\overline{\\rm MS}}=341$ MeV used to define the two-loop running coupling $\\hat g^2(T)$.","marker":"[43]"}],"fun_headline_variants":["Screening masses up to 160 GeV defy perturbative QCD","Higher-order terms persist in thermal QCD to 160 GeV","NLO not enough: QCD screening masses demand more up to 160 GeV","Hadronic screening at 160 GeV breaks perturbative predictions","Non-perturbative QCD masses exceed NLO fits across full range"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that restricting all simulations to the zero-topological-charge sector is safe down to $T\\simeq1$ GeV; if the probability of nonzero topology is not several orders of magnitude below the permille level, the extracted screening masses could carry an unquantified bias.","fun_headline_variants_meta":{"raw":{"variants":["Screening masses up to 160 GeV defy perturbative QCD","Higher-order terms persist in thermal QCD to 160 GeV","NLO not enough: QCD screening masses demand more up to 160 GeV","Hadronic screening at 160 GeV breaks perturbative predictions","Non-perturbative QCD masses exceed NLO fits across full range"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000208,"raw_usage":{"total_tokens":1526,"prompt_tokens":1192,"completion_tokens":334,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":808,"completion_tokens_details":{"reasoning_tokens":241}},"tokens_in":808,"tokens_out":334,"duration_ms":3869,"temperature":1.0,"reasoning_tokens":241,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T15:30:56.813278+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the topological-charge distribution on the largest spatial-volume lattices at $T\\simeq1$ GeV, where the paper's boxes have $LT$ between 20 and 50, and compare screening masses extracted with and without the $Q=0$ restriction; if configurations with $|Q|>0$ occur with probability near or above $10^{-3}$, the continuum screening masses would shift by more than their quoted permille accuracy.","supporting_citations":[{"cited_title":"Baryonic thermal screening mass at NLO","cited_arxiv_id":"2405.03975","evidence_quote":"It provides the next-to-leading-order baryonic screening-mass calculation in the effective theory, eq. (16), obtained from the three-body quantum-mechanical problem."}],"review_version":1}