{"id":"34d99591-a221-422f-b0ff-8b3b93d159bc","arxiv_id":"2501.19192","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A modified Grassmann tensor-network algorithm computes strong-coupling coefficients of the QCD partition function up to order beta^3, and fit ansatze extend the quark number density and chiral condensate beyond the expansion's validity.","lead":"The authors show how to compute quark number density and chiral condensate in two-dimensional QCD using a tensor-network method in the strong-coupling expansion, with coefficients up to third order in the gauge coupling. Fits to the coefficients extend the results to larger couplings and yield a coupling-dependent estimate of the critical chemical potential.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central order-separation claim of Sec. 4 is asserted but not demonstrated; all fitted mu_c(beta) results inherit this unverified prerequisite.","rationale":"The abstract and Sec. 4 make a strong algorithmic promise: direct computation of all strong-coupling coefficients up to order n_max with no higher-order contamination. This is what distinguishes the paper from a naive truncated plaquette expansion, and it is the premise for trusting Figs. 3-5. The reader's verdict already flags that this is not demonstrated, but lists the tanh ansatz (8)/(11) as the weakest assumption. In my reading, the order-separation claim is more load-bearing than the ansatz: even if the tanh form were exact, contaminated coefficients would invalidate the extracted mu_c. Conversely, if the order-separation claim is correct, the ansatz systematics remain a fit-model uncertainty that could be addressed by higher-order data. Thus the single most important check is to verify order purity independently. The proposed n_max=4 comparison is a direct falsification test: if lower-order coefficients shift when the truncation ceiling is raised, contamination is present; if they are stable, the central claim gains strong support. I therefore keep the conditional verdict: the paper is not ready for acceptance as a full claim, but the concern is addressable in the forthcoming publication or by releasing code and cross-checks.","tokens_in":6735,"tokens_out":9173,"duration_ms":95564,"concrete_test":"Run the modified GHOTRG with n_max=4 on the same 32x32 lattice, m=0.5, and D=135, and compare the extracted rho_0,...,rho_3(mu) and <bar psi psi>_0,...,_3(mu) with the n_max=3 results. If any of the first four coefficients changes beyond the estimated D-convergence tolerance, the claim of complete absence of higher-order contamination fails. Repeat the comparison at D=200 to confirm that any observed shift is not merely bond-dimension truncation error.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim (abstract; Sec. 4) is that the modified GHOTRG computes mu-dependent coefficients through order beta^3 'with complete absence of higher-order terms infiltrating the result.' This is the load-bearing assertion: if any term above order 3 leaks into the extracted rho_n(mu) or <bar psi psi>_n(mu), then the simultaneous tanh fit in Sec. 6 and the resulting mu_c(beta) in Fig. 5 are not strong-coupling coefficients at all. The manuscript does not supply a demonstration. Sec. 4 describes the order-grouping and separate SVD truncations only verbally and sends the proof and implementation details to ref. [4], a forthcoming publication. No cross-check against an independent evaluation of the truncated expansion is reported; the stated consistency among n_max<3 and n_max=3 runs is consistency within the same algorithm, not an external validation of the absence of contamination. Moreover, GHOTRG itself is approximate, with bond dimension D=135. The paper does not specify why the order separation survives SVD truncation with finite D exactly rather than only approximately, nor how truncation error in one order sector is prevented from leaking into another. Since every subsequent physical result—the error bands in Fig. 4 and the beta expansion of mu_c in Fig. 5—is downstream of this unverified step, the central claim is not yet established. The ansatz caveats in footnotes 6 and 7 are secondary; even a perfect fit to contaminated coefficients would produce a wrong mu_c.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript presents a Grassmann tensor-network treatment of two-dimensional lattice QCD with staggered quarks in the strong-coupling expansion at finite chemical potential. The Boltzmann factor is expanded in the gauge and fermion actions, gauge fields are integrated exactly, and the partition function coefficients in powers of beta are evaluated with a modified GHOTRG that is claimed to separate orders in beta exactly. Results are shown for the quark number density and chiral condensate up to order beta^3 on a 32x32 lattice, together with tanh-based fits that extrapolate the transition parameters beyond beta=0.1. The paper is a proceedings contribution; details of the modified algorithm are deferred to Ref. [4].","tokens_in":7107,"tokens_out":5509,"duration_ms":49427,"significance":"If the order-separation claim is correct, the paper offers a practical route to exact-in-beta strong-coupling coefficients for 2d QCD at finite mu, with gauge degrees of freedom integrated out before tensor renormalization and with no sign problem. The coefficients rho_n(mu) and <psi psi>_n(mu) are computed, not fitted, and the consistency of n_max=3 with lower orders is encouraging. The main caveats are that the 'complete absence' of higher-order contamination is only asserted, and that the physical extrapolations rely on an ad hoc tanh ansatz rather than a derived beta expansion. The strength of the computation is its direct, non-perturbative-in-Grassmann method; its weakness is the lack of independent validation of the central algorithmic claim within this manuscript.","major_comments":[{"comment":"The abstract and Section 4 state that the modified GHOTRG procedure computes the mu-dependent quark number density and chiral condensate 'up to order beta^3 with complete absence of higher-order terms infiltrating the result.' This is the load-bearing claim, but the manuscript does not demonstrate it. Section 4 describes the order-grouping and separate SVD truncations only verbally and defers the detailed proof and implementation to Ref. [4], which is 'in preparation'. Since GHOTRG with D=135 is an approximate algorithm, it is not self-evident that finite bond-dimension truncations performed in one order sector cannot mix different orders; the consistency check between n_max<3 and n_max=3 runs in Sec. 6.1 is internal to the same algorithm and does not exclude a common contamination. Consequently, the coefficients in Figs. 3 and 4, and all mu_c(beta) results in Fig. 5, inherit this unverified prerequisite. I ask for an independent validation, for example by comparing against a brute-force evaluation of the truncated strong-coupling expansion on a small lattice, or by demonstrating numerically that no terms of order beta^4 survive in rho_3(mu).","section":"Section 4 and abstract"},{"comment":"The 'valuable expansion in beta for the critical chemical potential' reported in Sec. 6.3 is not a derived strong-coupling quantity but an output of the tanh fit ansatz (8) with the power series (9). The fit parameters mu_c,rho,n and a_rho,n are introduced by hand, and the error bands in Fig. 4 and the extrapolation beyond beta=0.1 inherit the validity of this functional form. The paper does not test the ansatz against an alternative (e.g., the antisymmetric combination acknowledged in footnote 6), nor does it show how the extracted mu_c(beta) changes when n_max is increased from 3 to 4. Without such a stability check, the statement that 'the critical chemical potential remains quite accurate' for beta up to 1 is not supported.","section":"Section 6.1, Eqs. (8)-(9)"},{"comment":"All numerical results use a single lattice size (32x32) and a single bond dimension (D=135). No analysis is shown of the dependence of rho_n(mu) or <psi psi>_n(mu) on D or lattice volume, so the reader cannot tell whether the reported coefficients are converged with respect to the tensor-network truncation. This matters because the claim that the coefficients are free of higher-order contamination is tied to the exactness of the order separation; if the separation is only approximate at finite D, the truncation error in each sector has to be quantified.","section":"Section 6.1, Figs. 3-4"}],"minor_comments":[{"comment":"The choice of 1% of the maximal value of |rho_n(mu)| as the error for each data point is not motivated; the resulting covariance matrix and error bands in Figs. 4 and 5 are sensitive to this choice. Please state how this uncertainty estimate was calibrated.","section":"Section 6.1"},{"comment":"The introduction of the four special tensors and the 'impurity system' is not illustrated; a figure or explicit tensor diagram would help the reader follow the translational-invariance reduction.","section":"Section 5"},{"comment":"The claim of 'excellent agreement' between tensor data and the simultaneous fit would be easier to assess with a residual plot or a quantitative goodness-of-fit statistic, especially because the error bars are assigned rather than estimated from the tensor computation.","section":"Section 6.1, Fig. 3"}],"recommendation":"major_revision","confidential_remarks":"For the editor: this is a proceedings-style paper whose central methodological claim is outsourced to Ref. [4]. If the journal allows such references, the paper may be acceptable as a short contribution, but the published version should clearly flag the conditional status of the order-separation claim. I would not reject on the basis of the fit ansatz, since the authors are transparent about its heuristic nature."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper is a proceedings write-up of a plausible tensor-network method for 2D QCD in the strong-coupling expansion. What's genuinely new is a modified GHOTRG that groups tensor indices by plaquette occupation and truncates within those groups, so that coefficients of the beta expansion are computed directly up to O(beta^3) rather than by fitting noisy beta-dependent data. That is a real algorithmic step beyond the same group's earlier infinite-coupling work, and the coefficients for rho_n(mu) and <psibar psi>_n(mu) are new. The simultaneous tanh fit that converts those coefficients into a beta expansion for mu_c is also new, and the authors are honest that it is a fit, not a derived quantity.\n\nWhat is solid: the paper is fairly clear for a proceedings; the agreement between n_max=3 and lower-order runs is encouraging, and the fits in Fig. 3 look good. The footnotes even flag symmetric-ansatz improvements, which is honest.\n\nThe soft spot is load-bearing. The abstract and Sec. 4 claim 'complete absence of higher-order terms infiltrating the result,' but the mechanism is only described verbally and the proof is deferred to a forthcoming paper [4]. GHOTRG itself is approximate (D=135), and the paper never explains why order separation survives SVD truncation sector by sector, rather than only approximately. Without an independent cross-check (e.g., against a direct contraction on a small lattice or an exact small-volume calculation), the claim is unverified. And everything downstream—the error bands in Fig. 4 and the mu_c(beta) expansion in Fig. 5—is built on that step. If contamination occurs, the coefficients are not strong-coupling coefficients, and the fits inherit the same problem.\n\nThe beta>0.1 extrapolation is also speculative, since the underlying expansion breaks down there and the tanh form is an ansatz. The authors acknowledge this, so it is a minor-to-moderate issue compared to the order-separation gap.\n\nThis is for people working on tensor-network approaches to lattice QCD at finite density. They will find the method worth examining, provided the substantiating paper exists.\n\nMy recommendation: the idea deserves a serious referee, but the current manuscript is not sufficient to establish the central claim. I would send it to review with instructions to the authors to either include the order-separation demonstration or provide a numerical cross-check against an independent evaluation of the truncated expansion. A proceedings publication can stand as a preliminary report, but the main result should not be cited as established until that proof appears.","headline":"Plausible algorithmic advance for strong-coupling tensor networks, but the central order-separation claim is asserted, not yet demonstrated.","tokens_in":7552,"tokens_out":3190,"would_cite":true,"duration_ms":31881,"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":"This paper computes two-dimensional QCD strong-coupling coefficients through order beta^3 with no higher-order contamination, then fits them to extract the critical chemical potential's beta expansion.","keywords":["strong-coupling expansion","lattice QCD","tensor network","Grassmann variables","higher-order tensor renormalization group","chemical potential","chiral condensate","quark number density"],"falsifier":"Compute the $\\beta^4$ coefficient of the quark number density with the same order-separating method: the claimed absence of contamination would be falsified if the $\\beta^3$ coefficient changes when $n_{\\max}$ is raised from 3 to 4, and the $\\tanh$ extrapolation would be falsified if the cubic-parameter fit fails to predict the $\\beta^4$ coefficient. A complementary check would compare the direct tensor coefficients at $\\beta\\le0.1$ with independent Monte Carlo or dual-representation data at the same lattice size and quark mass.","tokens_in":6559,"feed_emoji":"🧮","tokens_out":11946,"duration_ms":99430,"temperature":0.7,"pith_summary":"The paper sets out to show that two-dimensional lattice QCD with staggered quarks at nonzero chemical potential can be rewritten as a Grassmann tensor network, and that a modified higher-order tensor renormalization group computes the strong-coupling expansion of the partition function order by order. Its central claim is that the quark number density and chiral condensate can be obtained directly through order $\\beta^3$, with no higher-order terms leaking into those coefficients. If true, the reported tensor data are exact strong-coupling coefficients, not contaminated approximations, and the fitted $\\beta$-dependence of the critical chemical potential is a quantitative prediction for where the transitions occur. This matters because the strong-coupling expansion is one of the few controlled routes into the finite-density region of QCD where standard Monte Carlo sampling fails.","feed_headline":"2D QCD strong-coupling coefficients pinned down to order beta^3","feed_subtitle":"Modified Grassmann tensor renormalization stops higher-order terms from leaking into the beta^3 coefficients.","key_machinery":"The load-bearing object is the Grassmann higher-order tensor renormalization group (GHOTRG), a tensor-network contraction scheme that compresses coarse-grained tensors by singular-value decompositions while tracking Grassmann signs through auxiliary link variables. The paper adds a modified truncation: link indices are grouped according to the occupation of adjacent plaquettes, and separate SVD truncations are performed on each group so that contributions of different orders in $\\beta$ are never mixed during coarsening. Plaquette contractions into links and links into sites are handled by site-dependent tensors, which is what allows all coefficients up to a given order to be computed simultaneously. Before any of this, all $SU(N_c)$ gauge integrals are performed exactly using generalized Weingarten functions, which eliminates all color degrees of freedom and keeps the initial bond dimension small.","core_discovery":"The central claim is that the modified Grassmann tensor renormalization procedure computes the strong-coupling coefficients of the quark number density and chiral condensate through order $\\beta^3$ with complete absence of higher-order terms infiltrating the result. Working at expansion order $n_{\\max}=3$ on a $32\\times32$ lattice with quark mass $m=0.5$ and bond dimension $D=135$, the paper reports coefficients $\\rho_n(\\mu)$ and $\\langle \\bar\\psi\\psi\\rangle_n(\\mu)$ for $n=0,\\ldots,3$ that are stable when the truncation order is raised and consistent between neighboring bond dimensions. All coefficients are fitted simultaneously to a Fermi-Dirac-like $\\tanh$ ansatz for the quark number density and an analogous $\\tanh$ form for the chiral condensate, with the transition parameters expanded as power series in $\\beta$ through third order. The fit reproduces the tensor data within the assumed errors, and the extracted critical chemical potentials for the two transitions agree within errors, with second- and third-order terms giving only small corrections. The paper's claim is that these fits remain reliable beyond $\\beta=0.1$, where the direct expansion itself becomes unphysical.","pith_inferences":["Beyond the paper, the same order-separating Grassmann tensor construction should transfer to other lattice gauge theories whose Boltzmann weight can be expanded and whose gauge integrals admit a Weingarten-type solution, making the method a general template for finite-density strong-coupling computations.","Beyond the paper, watching the fitted transition sharpness $a_\\rho(\\beta)$ and $a_{cc}(\\beta)$ as $\\beta$ grows could indicate whether the crossover sharpens into a genuine phase transition, a question the paper does not address.","Beyond the paper, the contamination-free coefficients from this small two-dimensional system are natural benchmark data for sign-problem-avoiding algorithms, since they carry no Monte Carlo statistical error or phase ambiguity."],"forward_implications":["The coefficients $\\rho_n(\\mu)$ and $\\langle \\bar\\psi\\psi\\rangle_n(\\mu)$ for $n\\le3$ are strong-coupling reference data at finite chemical potential that other computational approaches can be checked against.","The fitted expansions give a quantitative prediction that the critical chemical potential increases with the coupling $\\beta$ in both the number-density and chiral transitions, with the two estimates agreeing within errors.","The method is stated to generalize straightforwardly to an arbitrary number of staggered quark flavors, extending the same order-separated tensor construction beyond the single-flavor case.","Because the direct expansion fails for $\\beta>0.1$ while the $\\tanh$ fits remain stable, the fit-based extrapolation is the paper's route to accessing larger couplings from strong-coupling data."],"supporting_citations":[{"why":"supplies the Grassmann HOTRG approach for two-dimensional strong-coupling QCD that this work extends beyond infinite coupling","marker":"[1]"},{"why":"contains the full derivation of the modified order-separating GHOTRG procedure whose absence-of-contamination property is the central claim","marker":"[4]"},{"why":"introduces the higher-order tensor renormalization group contraction and truncation scheme on which the algorithm is built","marker":"[5]"},{"why":"provides the multilinear singular value decomposition used in the HOTRG truncation step","marker":"[6]"},{"why":"provides the SU(N) polynomial integral formula used to integrate out all gauge fields exactly","marker":"[7]"},{"why":"offers an independent efficient solution of the same gauge integrals, supporting the integration step","marker":"[8]"},{"why":"introduces the Grassmann network construction used to handle Grassmann signs and produce Grassmann link variables","marker":"[9]"}],"fun_headline_variants":["Grassmann tensor nets clean beta^3 QCD coefficients","Beta^3 strong-coupling QCD coefficients with zero leakage","Tensor network purifies 2D QCD strong-coupling to beta^3","Clean beta^3 coefficients for strong-coupling 2D QCD","Grassmann renormalization locks QCD beta^3 terms exactly"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusions rest on the assumption that the quark number density and chiral condensate are well described, over the fitted range, by the hyperbolic-tangent forms (8) and (11) with their parameters expanded as power series in $\\beta$ through third order; if that functional form is wrong, the extracted critical chemical potential and its $\\beta$ expansion are artifacts of the fit even though the underlying tensor coefficients may be correct.","fun_headline_variants_meta":{"raw":{"variants":["Grassmann tensor nets clean beta^3 QCD coefficients","Beta^3 strong-coupling QCD coefficients with zero leakage","Tensor network purifies 2D QCD strong-coupling to beta^3","Clean beta^3 coefficients for strong-coupling 2D QCD","Grassmann renormalization locks QCD beta^3 terms exactly"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000265,"raw_usage":{"total_tokens":1601,"prompt_tokens":931,"completion_tokens":670,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":547,"completion_tokens_details":{"reasoning_tokens":576}},"tokens_in":547,"tokens_out":670,"duration_ms":6218,"temperature":1.0,"reasoning_tokens":576,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T20:57:36.098502+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the $\\beta^4$ coefficient of the quark number density with the same order-separating method: the claimed absence of contamination would be falsified if the $\\beta^3$ coefficient changes when $n_{\\max}$ is raised from 3 to 4, and the $\\tanh$ extrapolation would be falsified if the cubic-parameter fit fails to predict the $\\beta^4$ coefficient. A complementary check would compare the direct tensor coefficients at $\\beta\\le0.1$ with independent Monte Carlo or dual-representation data at the same lattice size and quark mass.","supporting_citations":[{"cited_title":"Milde, Tensor networks for four-dimensional strong-coupling QCD , Master's thesis, University of Regensburg, 2023","cited_arxiv_id":null,"evidence_quote":"contains the full derivation of the modified order-separating GHOTRG procedure whose absence-of-contamination property is the central claim"},{"cited_title":"Samberger, J","cited_arxiv_id":null,"evidence_quote":"introduces the higher-order tensor renormalization group contraction and truncation scheme on which the algorithm is built"}],"review_version":1}