{"id":"a17251e5-6343-4ca3-aa3b-4742e0ae20b3","arxiv_id":"2502.01396","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"The interface tension and latent heat of the SU(N) deconfinement transition scale as N^2 in the continuum limit, with sigma/Tc^3 = 0.0189(11) N^2 - 0.190(19) and L/Tc^4 = 0.354(2) N^2 - 1.65(10).","lead":"Lattice simulations of pure SU(N) gauge theories show that both the confined-deconfined interface tension and the latent heat grow as the square of the number of colors N, confirming large-N counting. The paper gives continuum-extrapolated values for N up to 10 using the Moore-Turok capillary wave method.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The N^2 interface-tension result rests on a two-point linear a^2 continuum extrapolation with no systematic error; until N_t=10 or 12 data are added, Eq. (9) is not robust.","rationale":"I agree with the reader's weakest assumption. The paper's method is well-motivated and the data collapse in Fig. 4 is reassuring, but the headline numerical result is a continuum extrapolation built from two points per N. This is the classic place where lattice results fail: a linear a^2 fit through two points cannot detect curvature, and the absence of a quoted systematic error means the error bars in Eq. (9) are statistical only. The latent-heat result suffers similarly from the saturated interpolation in Eq. (12) and the exclusion of N_t=5. I would keep the verdict at CONDITIONAL: the qualitative conclusion that both quantities scale as N^2 is plausible and consistent with large-N expectations, but the quantitative coefficient and reference values are not established beyond reasonable doubt. The proposed test is direct and would settle whether the continuum extrapolation actually holds.","tokens_in":7418,"tokens_out":6665,"duration_ms":62383,"concrete_test":"Run the same Moore-Turok measurement at N_t=10 (and preferably N_t=12) for N=5, 8, and 10, using the same lattice geometries and smearing/kernel analysis as in Table 1. Then fit sigma/Tc^3 versus a^2 with (i) the published linear form through N_t=6,8,10, and (ii) a quadratic form including an O(a^4) term. If the inferred continuum values shift by more than the current N_t=6/8 extrapolation error, or if the linear fit has chi^2/dof greater than about 2, Eq. (9) needs revision and the central N^2 slope is not yet secure.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing condition for the central claim is that Eq. (9) is a continuum-limit statement. For each N, the continuum sigma/Tc^3 values in Table 1 come from exactly two lattice spacings, N_t=6 and N_t=8, extrapolated linearly in a^2. This suppresses any O(a^4), logarithmic, or non-linear discretization term that could tilt the intercept. With only two points, the extrapolation has zero internal degrees of freedom and no systematic error is assigned. The fitted coefficient 0.0189(11) is then determined by only four (or three, if N=4 is excluded) continuum values; if, for example, the N=5 continuum sigma shifted upward by about 8% and the N=8 value downward by a similar fraction, the fitted slope would move by several quoted errors. The paper also does not report the selected k-range or fit form for the k->0 extrapolation at each smearing level, so the stability shown in Fig. 4 does not by itself bound this systematic. The latent-heat side has a related but distinct vulnerability: Eq. (12) uses a three-parameter interpolation that saturates the beta_c(N_t) data, and the N_t=5 point is excluded from the continuum fit because it deviates; this makes those continuum values model-dependent too. Because Eqs. (9) and (13) are both intended as reference large-N results, the two-point sigma extrapolation is the least controlled element supporting the paper's headline.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings paper reports a high-precision lattice study of the confined-deconfined interface tension and latent heat in pure SU(N) gauge theory at large N. Using the Moore-Turok mixed-phase method, the authors restrict the average Polyakov loop to a narrow window in the coexistence region, determine beta_c by demanding a flat restricted distribution, and extract the interface tension from the capillary-wave spectrum of the interface after applying a semi-analytic smearing-kernel correction. Simulations are performed for N = 4, 5, 8, 10 and 16 on lattices with N_t from 5 to 8. The paper finds sigma/T_c^3 = 0.0189(11) N^2 - 0.190(19) in the continuum limit and L/T_c^4 = 0.354(2) N^2 - 1.65(10) for N >= 5, and concludes that both quantities scale as N^2 at large N.","tokens_in":7783,"tokens_out":4482,"duration_ms":42856,"significance":"If the result holds, this paper provides the first precise lattice determination of the interface tension in the large-N limit and sharpens the existing latent-heat determination by more than an order of magnitude. The methodology has genuine strengths: the mixed-phase restriction avoids supercritical slowing down, the smearing-kernel correction collapses the Fourier spectrum onto a single curve, the beta_c determination is visually and statistically robust, and the latent-heat result is compatible with the older SU(N) determination of Lucini, Teper and Wenger while being much more precise. The use of independent observables for sigma and L avoids a circular N^2 fit. However, the central large-N coefficients currently rest on a two-point continuum extrapolation for the interface tension and on a saturated interpolation ansatz for the beta-function derivative, so the numerical claims are not yet as secure as the paper's wording suggests.","major_comments":[{"comment":"The continuum value of sigma/T_c^3 for each N is obtained from exactly two lattice spacings, N_t = 6 and N_t = 8, via a linear extrapolation in a^2. Such a two-point extrapolation has zero internal degrees of freedom, cannot detect O(a^4) or logarithmic corrections, and no systematic error is assigned to it. Since the fitted N^2 coefficient 0.0189(11) is driven by four continuum points, a correlated shift of a few percent in the N = 5 and N = 8 intercepts would move the slope by several quoted errors. Please add at least one more lattice spacing for at least one representative N, or provide a quantitative estimate of the truncation uncertainty and include it in the final error.","section":"Section 3, Table 1, Eq. (9)"},{"comment":"The k -> 0 extrapolation of the kernel-corrected Fourier spectrum is not documented in enough detail. The text does not state the fit ansatz (e.g., constant plus linear or quadratic term in k^2), the momentum range used, the number of Fourier modes, or the chi^2 per degree of freedom. Only a single representative case (SU(16) at N_t = 6) is shown. Because sigma is the inverse of the intercept, a small systematic tilt in this fit propagates directly into every continuum point in Table 1. Please report the fit range, fit form, goodness of fit, and a table showing the stability of the extracted sigma across smearing levels for each N.","section":"Section 3, Fig. 4"},{"comment":"The derivative d beta / d ln a used in Eq. (11) is obtained from the three-parameter interpolation ansatz of Eq. (12), which the authors themselves state 'saturates the degrees of freedom of the fit.' With only four N_t values per N, this leaves no residual degrees of freedom, and the N_t = 5 point is then excluded from the continuum limit because it deviates. The continuum L/T_c^4 values and hence the N^2 coefficient in Eq. (13) are therefore partly dependent on the chosen functional form. Please add a stability test with a different interpolation ansatz or with the N_t = 5 point included, and propagate any resulting shift as a systematic uncertainty.","section":"Section 4, Eq. (12)"}],"minor_comments":[{"comment":"The rendered text contains numerous missing spaces (e.g., 'Wepresentresultsfrom...', 'Theresultsoftheextrapolationgivestheinverseofsigma'), which makes the manuscript difficult to read. These should be fixed in the final version.","section":"Throughout"},{"comment":"The phrase 'we observe unambiguously' is stronger than the current two-point continuum extrapolation for sigma supports; a wording such as 'consistent with N^2 scaling' would better match the evidence presented.","section":"Abstract and Conclusion"},{"comment":"The large-N latent-heat fit is stated to apply for N >= 5, but the text does not explicitly list which continuum points enter the fit (N = 5, 8, 10) and whether N = 4 is excluded from both the sigma and latent-heat fits. Please state this explicitly in a caption or in the text.","section":"Fig. 5, Eq. (13)"},{"comment":"The SU(3) comparison points from refs. [20] and [21] are shown in the figures but not discussed in the text. Since they provide an important external consistency check, a brief sentence describing the comparison would be useful.","section":"Section 3, Fig. 5"}],"recommendation":"major_revision","confidential_remarks":"The paper is a proceedings contribution, and the missing details may exist in a longer companion paper. I would encourage the editor to request either a supplementary file or an explicit statement of where the fit ranges, fit forms, and systematic checks are documented, because without them the central N^2 coefficients cannot be independently verified."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: the N^2 scaling of both the interface tension and latent heat is probably right, and this is the first continuum-extrapolated interface tension for pure SU(N). But the continuum limit that drives Eq. (9) rests on exactly two lattice spacings, so treat the quoted slope as preliminary until there are more N_t values or a real systematic error.\n\nWhat is genuinely new: applying the Moore-Turok capillary method to SU(4,5,8,10) at N_t=6 and 8, plus a single SU(16) point, and getting a continuum limit for sigma/Tc^3. The semi-analytic smearing-kernel correction is a real plus: the collapse in Fig. 4 is convincing, and it shows the k->0 extrapolation is stable over a range of smearing levels. The beta_c determination from the flat restricted Polyakov-line distribution is also clean and robust. The latent heat result is basically a refined version of ref. [6], with smaller errors and compatible central values, which is still useful.\n\nThe soft spots are real but not fatal. The interface tension continuum extrapolation uses only N_t=6 and 8, linearly in a^2, with zero internal degrees of freedom and no assigned systematic from O(a^4) or log corrections. That is the main load-bearing weakness. If the N=5 or N=8 continuum values shift by a few percent, the fitted N^2 coefficient moves by more than its quoted error. The paper also does not report the selected k-range or fit form for the k->0 extrapolation, so the stability shown in Fig. 4 is not fully auditable. On the latent heat side, Eq. (12) uses a three-parameter interpolation that saturates the beta_c data, and N_t=5 is excluded post hoc from the continuum fit; that makes those values somewhat model-dependent. None of this overturns the central qualitative claim, and the external SU(3) point from ref. [20] is consistent with the trend, so the overall picture holds.\n\nCitation pattern looks fine: ref. [6] is the obvious predecessor, and the comparison with it is honest. The paper ships no code but refers to the HILA framework; the method is described well enough to reproduce with effort.\n\nWho is this for: lattice gauge theorists working on large-N thermodynamics, and people using holographic or effective models of deconfinement who need reference values. It deserves a serious referee, not a desk reject. The referee should push for N_t=10 or 12 data, or at least an explicit systematic error on the a^2 extrapolation, and for reporting the k-range and fit form. If the authors add that, Eq. (9) becomes a much stronger reference.","headline":"First continuum extrapolation of SU(N) interface tension, with plausible N^2 scaling, but the sigma continuum limit rests on two lattice spacings and should be treated as preliminary.","tokens_in":8281,"tokens_out":1670,"would_cite":true,"duration_ms":17276,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Pure SU(N) gauge theory shows interface tension and latent heat scaling as N^2 at large N.","keywords":["SU(N) gauge theory","large-N limit","confined-deconfined transition","interface tension","latent heat","lattice gauge theory","Polyakov loop","capillary waves"],"falsifier":"Take data at a third lattice spacing, such as $N_t=10$, at one $N$ (for instance $N=8$) and check whether $\\sigma/T_c^3$ continues to follow the linear-in-$a^2$ line used here; the current fit predicts the continuum value $1.004(94)$ from $N_t=6,8$, so an $N_t=10$ point that falls off that line, or a continuum value that disagrees with the $N^2$ fit, would falsify the scaling.","tokens_in":7191,"feed_emoji":"⚛️","tokens_out":6134,"duration_ms":51065,"temperature":0.7,"pith_summary":"Pure SU(N) gauge theories with N colors undergo a first-order confinement-deconfinement transition. This paper reports lattice measurements showing that the interface tension between the confined and deconfined phases, and the latent heat released in the transition, both grow as $N^{2}$ at large N. The result matters because the interface tension controls bubble nucleation and the dynamics of the transition, and was previously poorly determined. The paper extracts it from the thermal fluctuation spectrum of the phase interface using a constrained mixed-phase simulation, and takes a continuum limit.","feed_headline":"Confined-deconfined interface tension scales as N^2","feed_subtitle":"Lattice data for N=4 to 16 pin down sigma/Tc^3 = 0.0189 N^2 - 0.190.","key_machinery":"The central object is the capillary-wave spectrum of the phase interface, Eq. (1): for a nearly flat interface of area $L^2$, the Fourier modes of its height satisfy $\\langle |\\hat{z}(n_x,n_y)|^2\\rangle = T/(4\\pi^2 \\sigma (n_x^2+n_y^2))$, so the interface tension $\\sigma$ is the slope of $1/\\langle|\\hat{z}|^2\\rangle$ versus $k_x^2+k_y^2$. The argument works because the simulation is held in the mixed phase, so the interface can be located by thresholding the smeared Polyakov loop, and the smearing distortion is removed by dividing by the known Fourier transform of the smearing kernel. This turns the measurement of a strongly suppressed mixed-phase probability into a measurement of equilibrium height fluctuations, which are not suppressed and allow large volumes.","core_discovery":"In the continuum limit, the confined-deconfined interface tension and the latent heat of pure SU(N) gauge theory at large N are described by $\\sigma/T_c^3 = 0.0189(11) N^2 - 0.190(19)$ and $L/T_c^4 = 0.354(2) N^2 - 1.65(10)$ for $N \\ge 5$. The paper establishes this by simulating $N=4, 5, 8, 10$ and 16 on lattices with inverse temperatures $N_t = 5\\ldots 8$ (and $N_t=6$ for $N=16$), constraining the real part of the average Polyakov loop to stay in the mixed-phase region so that two interfaces coexist. The interface tension is read off from the long-wavelength spectrum of interface height fluctuations, after undoing the effect of the smearing kernel analytically. The latent heat comes from the plaquette discontinuity at the critical coupling, using the measured critical couplings to evaluate the $\\beta$ function. Both quantities extrapolate linearly in the squared lattice spacing to the continuum, and the continuum values follow the $N^2 + \\mathrm{const}$ form.","pith_inferences":["The exponentially small mixed-phase probability implied by $\\sigma \\propto N^2$ means that in strongly first-order large-$N$ theories bubble nucleation proceeds through a very suppressed channel; the measured tension is the input needed to quantify supercooling and nucleation rates, which the paper does not compute.","A direct test of the large-$N$ prediction would be a continuum extrapolation at $N=16$ from a second lattice spacing; the fit predicts $\\sigma/T_c^3 \\approx 4.65$ while the single $N_t=6$ value is $5.36(9)$, so the extrapolation must move down by roughly 0.7 if the linear $a^2$ assumption holds.","The same fluctuation-spectrum technique could be applied to other order parameters, for example in theories with matter multiplets, where the interface tension is less constrained; nothing in the method ties it to pure glue."],"forward_implications":["The fitted value $\\sigma/T_c^3 = 0.0189(11) N^2 - 0.190(19)$ gives a reference interface tension for all $N$, replacing the previously uncertain estimates.","The latent heat fit $L/T_c^4 = 0.354(2) N^2 - 1.65(10)$ reduces the error on the large-$N$ transition strength by an order of magnitude compared with earlier work.","Both scalings confirm the large-$N$ expectation that the deconfinement transition becomes stronger quadratically with $N$, with finite-$N$ corrections encoded in the small constant terms.","The constrained mixed-phase method is shown to work up to $N=16$ and avoids the supercritical slowing down of multicanonical methods, so it can be applied to transitions where the mixed-phase probability is extremely suppressed."],"supporting_citations":[{"why":"Introduces the constrained mixed-phase method and the interface fluctuation spectrum used to extract the interface tension.","marker":"[1]"},{"why":"Provides the earlier large-N latent heat and interface tension estimates that this work compares with and improves.","marker":"[6]"},{"why":"Shows the same constrained simulation technique applied to electroweak bubble nucleation, establishing its reliability.","marker":"[19]"},{"why":"Supplies the SU(3) interface tension value used as a comparison point in the N-dependence plot.","marker":"[20]"},{"why":"Supplies the precise SU(3) continuum latent heat used for comparison.","marker":"[21]"},{"why":"Provides previous SU(N_c) deconfinement studies that give context for the critical temperature and latent heat.","marker":"[5]"}],"fun_headline_variants":["SU(N) gauge: interface tension and latent heat scale as N^2","Large N confinement: surface tension grows like N^2","N^2 scaling confirmed for SU(N) interface tension","Lattice results: sigma/Tc^3 = 0.0189 N^2 - 0.190"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The result depends on the continuum extrapolation of the interface tension, which is a straight line in the squared lattice spacing drawn through only two lattice spacings ($N_t=6$ and $N_t=8$); if higher-order corrections in the spacing are not negligible or vary with $N$, the quoted $N^2$ coefficient would shift.","fun_headline_variants_meta":{"raw":{"variants":["SU(N) gauge: interface tension and latent heat scale as N^2","Large N confinement: surface tension grows like N^2","N^2 scaling confirmed for SU(N) interface tension","Lattice results: sigma/Tc^3 = 0.0189 N^2 - 0.190"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000643,"raw_usage":{"total_tokens":2917,"prompt_tokens":868,"completion_tokens":2049,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":484,"completion_tokens_details":{"reasoning_tokens":1966}},"tokens_in":484,"tokens_out":2049,"duration_ms":12649,"temperature":1.0,"reasoning_tokens":1966,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T15:27:19.976243+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take data at a third lattice spacing, such as $N_t=10$, at one $N$ (for instance $N=8$) and check whether $\\sigma/T_c^3$ continues to follow the linear-in-$a^2$ line used here; the current fit predicts the continuum value $1.004(94)$ from $N_t=6,8$, so an $N_t=10$ point that falls off that line, or a continuum value that disagrees with the $N^2$ fit, would falsify the scaling.","supporting_citations":[{"cited_title":"Classical Field Dynamics of the Electroweak Phase Transition","cited_arxiv_id":"hep-ph/9608350","evidence_quote":"Introduces the constrained mixed-phase method and the interface fluctuation spectrum used to extract the interface tension."}],"review_version":1}