{"id":"9a7583ab-278e-4808-91c3-bfde51ef30f6","arxiv_id":"1908.06291","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The deviation of the critical square lattice Ising model from the Ising CFT is governed by a spin-4 T^2 perturbation with coupling g = (2π^2/7N^2)(1 - e^{-8iu}), verified exactly at order 1/N^2 for multiple state families.","lead":"This paper identifies the precise irrelevant correction that makes the finite-size critical square lattice Ising spectrum differ from the continuum conformal field theory (CFT) limit. It derives the coupling constant from the lattice anisotropy and verifies it against exact transfer matrix eigenvalues for several families of states.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Narrow spectrum identities are solid; the global uniqueness of T^2 as the complete leading irrelevant perturbation is not yet established.","rationale":"The reader's CONDITIONAL verdict is appropriate. The paper derives the coupling constant with no free parameters and checks it against three infinite state families plus a degenerate pair, including a vanishing off-diagonal matrix element; these are exact algebraic identities, not numerics. The weak point is genuinely the uniqueness and extrapolation step: all tested states are one- or two-fermion states, and possible dimension-4 competitors such as T\\bar T are not analyzed. The factorized form of the lattice expansion (2.16)-(2.17) strongly constrains any left-right mixing operator, but this is not exploited. The proposed four-fermion check would settle whether the integrable structure is additive as implicitly assumed. The off-by-one in (3.16)-(3.17) is a typographical error corrected by replacing p with p-1 and does not change the verified identities. Therefore the reader's verdict remains CONDITIONAL, and the stress-test does not require moving to a different verdict.","tokens_in":11499,"tokens_out":13378,"duration_ms":151058,"concrete_test":"Take the NS four-fermion state psi_{-5/2} psi_{-3/2} psi_{-1/2} \\bar psi_{-1/2} |0;0\\rangle, corresponding to sign flips mu_1 = mu_2 = mu_3 = -1 and \\bar mu_1 = -1 in the r = 1 sector. From (2.16)-(2.17), the lattice prediction for M log Lambda at order 1/N^3 is the sum of the three holomorphic single-flip terms for p = 1, 2, 3 plus the antiholomorphic term for p = 1. Independently compute the diagonal matrix element of g I_3 + \\bar g \\bar I_3 on this four-fermion state using (3.4) and (3.22). If it equals the lattice sum, additivity and completeness of T^2 + \\bar T^2 on this class are confirmed; if it differs by an interaction term, the abstract claim that the leading irrelevant perturbation is identified is not supported for the full spectrum.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Equations (3.12)-(3.13) fix g and \\bar g from the vacuum state, and (3.5)-(3.7) then reproduce the lattice 1/N^3 coefficients for the NS two-fermion family (Section 3.2), the Ramond family (Section 3.3), and two degenerate level-4 states (Section 3.4). The load-bearing gap is the converse direction: these families do not establish that H_int = g T^2 + \\bar g \\bar T^2 is the complete dimension-4 perturbation of the lattice Ising CFT. In particular, the spin-0 operator T\\bar T has the same scaling dimension and is never mentioned in Section 3. Its diagonal expectation on a state of holomorphic/antiholomorphic dimensions (Delta, \\bar Delta) generically contains a term proportional to (Delta - c/24)(\\bar Delta - c/24), which would introduce left-right cross terms p \\tilde p that are absent from the factorized lattice expressions (2.16)-(2.17). One can only conclude that on the tested states such additional operators either vanish or are absorbed into the fitted couplings. Moreover, every test uses at most two fermions; a four-fermion state would probe whether I_3 eigenvalues are simply additive, a property used implicitly by the factorized lattice side but not derived for the perturbed CFT side. Thus the narrow statement that the selected families match is solid, but the abstract's claim that 'the leading irrelevant perturbation ... is identified' requires the untested assumption that no other dimension-4 operator contributes to the full spectrum.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the critical square-lattice Ising model with periodic boundary conditions and identifies the leading irrelevant operator in the conformal-field-theory description of its finite-size spectrum. The author uses exact transfer-matrix eigenvalues, expands them in 1/N, and matches the subleading terms to a perturbed CFT with H_int = g T^2_cyl + \\bar g \\bar T^2_cyl. The coupling g is fixed from the vacuum energy shift and then used to predict 1/N^3 corrections for several families of states, all in exact agreement with the lattice expansion. The paper also relates the perturbing operators to higher integrals of motion.","tokens_in":11759,"tokens_out":15571,"duration_ms":145503,"significance":"If the identification is correct, the paper provides a precise, parameter-free bridge between the lattice Ising model and perturbed CFT at next-to-leading order, with exact analytic checks for infinite families of states. The main strengths are the exact lattice eigenvalue formulas, the explicit expression for g in terms of the anisotropy u, and the verification that a single coupling fixed from the vacuum reproduces the subleading corrections for the NS and Ramond families and for degenerate level-4 states. The matching identities are exact and the comparison to Ref. [25] provides an independent consistency check. However, the uniqueness of the T^2+\\bar T^2 perturbation is not established, and this is a load-bearing point for the abstract's central claim.","major_comments":[{"comment":"The perturbation ansatz omits the dimension-4 spin-0 operator T\\bar T, which lives in the identity module and is not forbidden by the square-lattice 90-degree rotational symmetry. The lattice expressions (2.16)-(2.17) and (3.14)-(3.18) show a factorized left-right structure with no p\\tilde p cross terms, whereas a nonzero T\\bar T coupling would contribute to the two-fermion states of Section 3.2 a state-dependent term of the form (p-1/2-c/24)(\\tilde p-1/2-c/24). To support the abstract's claim that the leading irrelevant perturbation has been identified, the author should either compute the T\\bar T matrix elements on the tested states and show that consistency with the lattice data forces its coefficient to vanish, or explicitly limit the claim to the T^2+\\bar T^2 perturbation and state that other dimension-4 operators are not excluded.","section":"Section 3, Eq. (3.1)"},{"comment":"All tested states contain at most two fermions per chirality. The lattice side uses additivity of single-flip contributions (Section 2.4), but the perturbed CFT side with H_int = g I3 + \\bar g \\bar I3 is not shown to have additive eigenvalues for states with more than two excitations. A check on a state with two left and two right fermions, for example ψ_{-p+1/2}ψ_{-q+1/2}\\bar ψ_{-\\tilde p+1/2}\\bar ψ_{-\\tilde q+1/2}|0\\rangle, would test this additivity and would further support the identification of the perturbation. This is a request for additional evidence rather than an identified error, but it is relevant to the strength of the central claim.","section":"Sections 3.2-3.4"}],"minor_comments":[{"comment":"The equation as printed uses M(1/2,p), but the state (3.16) has level p-1 above |1/2;1/2>. For p=1 the displayed identity fails; the correct form is g M(1/2,p-1) on the left-hand side, or equivalently p should be replaced by p-1 in the argument of M. This is a typographical error in the displayed equation, not in the surrounding derivation.","section":"Section 3.2, Eq. (3.17)"},{"comment":"The notation L^p_{-1} versus L^{p-1}_{-1} is easily confused in the rendering; please standardize the superscript and subscript placement so that the distinction between the number of L_{-1} insertions and the lattice label p is unambiguous.","section":"Section 3, Eqs. (3.5) and (3.16)"},{"comment":"The derivation of the factors relating g and gl is very terse; a short explanation of the geometric factor \\sqrt{2} and the phase factor e^{iθs} would help readers verify the consistency condition g = -(2π)^3/\\tilde N^2 gl.","section":"Section 3.1, comparison with Ref. [25]"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a concise and mostly careful analytic derivation. The main issue is the overstatement of uniqueness in the abstract: the absence of T\\bar T in the perturbation ansatz is not justified. This is fixable either by a direct calculation showing the T\\bar T coefficient must vanish or by softening the claim. The off-by-one typo in Eq. (3.17) and the ambiguous notation in Section 3 should also be corrected. If these points are addressed, the paper would be a solid contribution to the lattice/CFT correspondence at subleading order."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper does something real: it fixes the coupling constant for the T^2 + \\bar T^2 perturbation of the critical Ising CFT from the vacuum energy shift on the lattice, then shows that the same coupling reproduces the next-to-leading finite-size corrections for the NS two-fermion family, the Ramond family, and two degenerate level-4 states. The matching identities are exact and checkable, and the anisotropy dependence in (3.12)-(3.13) is new. The matrix elements come from published work, but the systematic lattice-side expansion and the comparison across these families is a solid piece of work.\n\nThe main soft spot is the interpretation. The abstract says \"the leading irrelevant perturbation is identified,\" but the author never mentions T\\bar T, the spin-0 dimension-4 operator in the identity module. On the tested states, a T\\bar T term would generically produce left-right cross terms like p \\tilde p, which are absent from the factorized lattice expressions. So the checks probably rule out T\\bar T, but the paper doesn't make that argument. The claim of uniqueness is therefore stronger than what is proven. The narrow statement—this particular perturbation with this coupling reproduces these families—is fully supported.\n\nThere is also a clear off-by-one typo in (3.16)-(3.17): the state is L^{p-1}_{-1} \\bar L^{\\tilde p-1}_{-1}|1/2;1/2>, not L^p_{-1}... The author's own p=1 case shows it. Cosmetic, but needs fixing. The notation in (2.18) is also muddled, and the abstract's \"1/N^2 corrections\" could use a pointer that the lattice eigenvalue shifts are 1/N^3 while the coupling is 1/N^2.\n\nWho should read it: people working on lattice/CFT correspondence, finite-size corrections, and integrable perturbations. It deserves a serious referee; I'd send it out. The referee should ask for the typo fix and a softening or proof of the uniqueness claim. The calculations are the paper's strength, and they hold up.","headline":"Solid exact matching for the T^2 perturbation of lattice Ising, but the claim to have identified the leading irrelevant perturbation outruns the evidence.","tokens_in":12287,"tokens_out":9268,"would_cite":true,"duration_ms":85901,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T40","82B20","82B23"],"pacs":["11.25.Hf","05.50.+q"],"model":"deepseek-v4-flash","headline":"One spin-4 irrelevant perturbation, with coupling $g = \\frac{2\\pi^2}{7N^2}(1-e^{-8iu})$, reproduces the next-to-leading $1/N^2$ spectrum of the critical square-lattice Ising model.","keywords":["Ising model","irrelevant perturbation","conformal field theory","finite-size corrections","transfer matrix","anisotropy parameter","higher integrals of motion","Virasoro algebra"],"falsifier":"Take a state with three sign flips on the left and three on the right in the transfer-matrix eigenvalues so that the total fermion-number constraint (2.6) is satisfied; its lattice $1/N^2$ shift is the sum of three copies of (2.16), while the perturbed CFT prediction is $g$ times the $T^2$ matrix element in the corresponding three-fermion state. If these two numbers differ, the single-operator identification fails outside the families checked in the paper; if they agree, the claim is strengthened toward a full-spectrum statement.","tokens_in":11265,"feed_emoji":"🧲","tokens_out":15229,"duration_ms":121396,"temperature":0.7,"pith_summary":"Critical square-lattice Ising model with periodic boundary conditions is not exactly conformal at finite lattice spacing: its transfer-matrix spectrum deviates from the conformal-field-theory prediction by computable $1/N^2$ corrections. The paper identifies the single irrelevant operator responsible for these deviations, the spin-4 combination $H_{\\text{int}} = g T^2 + \\bar g \\bar T^2$ built from the holomorphic and antiholomorphic stress tensors, and derives its coupling constant $g = \\frac{2\\pi^2}{7N^2}(1-e^{-8iu})$ from the lattice anisotropy parameter $u$. Substituting this $g$ into universal Virasoro matrix elements reproduces the lattice's next-to-leading energy shifts exactly for the vacuum, the energy-density descendants, the spin-field Ramond states, and a degenerate level-four pair. The point of the result is that a lattice artifact is not an unmodeled correction: it is one exactly parameterized irrelevant perturbation of the Ising CFT.","feed_headline":"Spin-4 operator sets lattice Ising's next-to-leading spectrum","feed_subtitle":"One complex number, fixed by anisotropy, reproduces the lattice's next-to-leading shifts.","key_machinery":"The load-bearing object is the regularized square of the stress tensor, $T^2_{\\text{cyl}}(\\zeta)=\\oint \\frac{T_{\\text{cyl}}(\\zeta')T_{\\text{cyl}}(\\zeta)\\,d\\zeta'}{2\\pi i(\\zeta'-\\zeta)}$, together with its antiholomorphic counterpart. In Virasoro modes this becomes $g\\left(2\\sum_{n\\ge1}L_{-n}L_n + L_0^2 - \\frac{c+2}{12}L_0 + \\frac{c(22+5c)}{2880}\\right)$ plus the conjugate term, which is the higher integral of motion $I_3$ of the Ising CFT. Because $I_3$ commutes with $I_1 = L_0 + \\bar L_0 - c/12$, the perturbation stays diagonal in the Virasoro basis; the paper can compare the exact large-$N$ expansion of the Baxter eigenvalue formulas, obtained by Euler-Maclaurin summation, with universal matrix elements of $I_3$ computed once for all states. The matching of the two independent computations fixes $g$ and then verifies it against infinite families of states.","core_discovery":"The paper's central discovery is that the leading deviation of the critical square-lattice Ising model from its continuum conformal limit is carried by a single irrelevant perturbation, $H_{\\text{int}} = g T^2_{\\text{cyl}} + \\bar g \\bar T^2_{\\text{cyl}}$, with $g = \\frac{2\\pi^2}{7N^2}\\left(1-e^{-8iu}\\right)$ and $\\bar g$ its complex conjugate; here $u$ parametrizes the lattice anisotropy and $N$ is the number of vertical lattice columns. The perturbing operator is the regularized square of the stress tensor and coincides with the higher integral of motion $I_3$ (plus $\\bar I_3$) of the Ising CFT, so it commutes with the unperturbed Hamiltonian and first-order perturbation theory is exact for the states considered. Inserting $g$ into the universal matrix elements $M(\\Delta,p)$ of $I_3$ reproduces, term by term, the $1/N^2$ corrections obtained from the exact eigenvalues of the transfer matrix for the vacuum $|0;0\\rangle$, the descendants $L^p_{-1}\\bar L^{\\tilde p}_{-1}|1/2;1/2\\rangle$, the Ramond states $\\psi_{-p}\\bar\\psi_{-\\tilde p}|1/16;1/16\\rangle$, and the degenerate pair $\\psi_{-7/2}\\psi_{-1/2}|0;0\\rangle$, $\\psi_{-5/2}\\psi_{-3/2}|0;0\\rangle$.","pith_inferences":["If the single-operator ansatz holds beyond the checked families, the entire $1/N$ expansion of the lattice model is a deformation of the Ising CFT by the commuting family of higher integrals of motion, with all coupling constants fixed by the anisotropy parameter $u$; this would make the finite-size spectrum a purely CFT-derived object.","The same matching strategy could be applied to other exactly solvable lattices: the spin of the perturbing field should reflect the lattice rotation symmetry, so one could predict the finite-size corrections before solving the model on that lattice.","A sharper test would be a degenerate multiplet in which $H_{\\text{int}}$ is not diagonal; exact lattice data would then decide whether the single-operator identification survives when mixing between descendant states is unavoidable.","One could also treat $g$ as a lattice observable by fitting the exact transfer-matrix eigenvalues at finite $N$ for states outside the proven families, effectively measuring the coupling numerically rather than deriving it from the vacuum shift."],"forward_implications":["The subleading finite-size spectrum of the critical toroidal Ising model becomes a CFT prediction with one input number, the coupling $g$; no lattice-specific fitting parameters remain.","For the vacuum, the energy-density descendants, the Ramond spin-field states, and the degenerate level-four pair, the perturbed CFT and the exact transfer matrix agree at order $1/N^2$ exactly, so the effective description is predictive rather than merely qualitative.","Since the perturbing operator is an integral of motion, the paper expects higher orders to be generated by the further conserved charges $I_5, I_7, \\ldots$, connecting the integrable structure of the lattice model to the integrable structure of the CFT.","In the isotropic case $u=\\pi/8$, one has $g=\\bar g=4\\pi^2/(7N^2)$, and the paper shows consistency with the previously known coupling of the unrotated lattice Ising model up to normalization, geometry, and a $\\pi/4$ rotation phase."],"supporting_citations":[{"why":"supplies the exact transfer-matrix eigenvalues of the periodic critical Ising model used as the lattice-side input.","marker":"[6]"},{"why":"gives the explicit factorized eigenvalue formulas (2.5)-(2.7) that are expanded at large N.","marker":"[7]"},{"why":"provides the conformal finite-size scaling prediction (2.10) used to identify q, the central charge, and the vacuum shift.","marker":"[12]"},{"why":"supports the universal free-energy/conformal-anomaly term entering the CFT prediction.","marker":"[13]"},{"why":"underpins the renormalization-group treatment of irrelevant perturbations near a two-dimensional critical point.","marker":"[17]"},{"why":"introduces the integrable structure of CFT in which the T^2 operator is realized as the higher integral of motion I3.","marker":"[20]"},{"why":"supplies the universal matrix-element formula M(Delta,p) for I3 that the paper uses to test perturbation theory.","marker":"[22]"},{"why":"provides the previously obtained coupling constant for the unrotated lattice, used as a consistency check in the isotropic case.","marker":"[25]"}],"fun_headline_variants":["Single coupling reproduces Ising lattice's 1/N^2 corrections","Exact lattice-CFT agreement from one irrelevant operator","Anisotropy fixed: one coupling sets lattice Ising's CFT deviation","Leading irrelevant operator controls lattice-CFT spectrum shift"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes that no other small-scale lattice effect contributes to the next-to-leading energy shifts except the spin-4 $T^2$ perturbation and its conjugate; the paper shows this is enough for the families it checks but does not prove it is the only possibility.","fun_headline_variants_meta":{"raw":{"variants":["Single coupling reproduces Ising lattice's 1/N^2 corrections","Exact lattice-CFT agreement from one irrelevant operator","Anisotropy fixed: one coupling sets lattice Ising's CFT deviation","Leading irrelevant operator controls lattice-CFT spectrum shift"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00082,"raw_usage":{"total_tokens":3589,"prompt_tokens":941,"completion_tokens":2648,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":557,"completion_tokens_details":{"reasoning_tokens":2577}},"tokens_in":557,"tokens_out":2648,"duration_ms":17564,"temperature":1.0,"reasoning_tokens":2577,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T12:51:26.354611+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a state with three sign flips on the left and three on the right in the transfer-matrix eigenvalues so that the total fermion-number constraint (2.6) is satisfied; its lattice $1/N^2$ shift is the sum of three copies of (2.16), while the perturbed CFT prediction is $g$ times the $T^2$ matrix element in the corresponding three-fermion state. If these two numbers differ, the single-operator identification fails outside the families checked in the paper; if they agree, the claim is strengthened toward a full-spectrum statement.","supporting_citations":[{"cited_title":"Baxter, Exactly Solved Models in Statistical Mechanics","cited_arxiv_id":null,"evidence_quote":"supplies the exact transfer-matrix eigenvalues of the periodic critical Ising model used as the lattice-side input."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"gives the explicit factorized eigenvalue formulas (2.5)-(2.7) that are expanded at large N."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the conformal finite-size scaling prediction (2.10) used to identify q, the central charge, and the vacuum shift."},{"cited_title":"Affleck, Universal term in the free energy at a critical point and the conformal anomaly, Phys","cited_arxiv_id":null,"evidence_quote":"supports the universal free-energy/conformal-anomaly term entering the CFT prediction."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"underpins the renormalization-group treatment of irrelevant perturbations near a two-dimensional critical point."},{"cited_title":"Reinicke, Analytical and non-analytical corrections to finite-size scaling , Journal of Physics A: Mathematical and General 20 (oct, 1987) 5325–5333","cited_arxiv_id":null,"evidence_quote":"supplies the universal matrix-element formula M(Delta,p) for I3 that the paper uses to test perturbation theory."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the previously obtained coupling constant for the unrotated lattice, used as a consistency check in the isotropic case."}],"review_version":1}