{"id":"8f77e01d-00c4-4652-81f8-442350b659a0","arxiv_id":"2512.20009","paper_version":2,"verdict":"CONDITIONAL","confidence":"LOW","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":1,"one_line_summary":"Closed-form Gershgorin-based upper bounds on the superconducting Tc in the γ-model improve on prior literature, while interlacing-based lower bounds reproduce known results at small truncation size.","lead":"Near quantum critical points, the γ-model captures electrons paired by singular boson fluctuations. This paper proves a new, tighter closed-form ceiling on the superconducting transition temperature for that model.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Upper bound Eq. 44 is not rigorously established for 0<γ≤1: the θ_m→0 boundary condition is imposed without proof and the N→∞ Gershgorin limit is asserted, so the advertised 'any γ>0' rigor is conditional.","rationale":"The reader's verdict is CONDITIONAL, and my stress-test agrees that the paper should not be accepted as a fully rigorous derivation of Eq. 44 for all γ>0 without closing the γ≤1 boundary-condition gap and the N→∞ Gershgorin step. I do not find a concrete algebraic error in the appendix's p=1/2 inequality; the identity used there is valid after symmetrization, and the finite-section-to-infinite-operator passage can likely be repaired by noting that finite Gershgorin positivity for all N implies positivity on finite support. However, the authors themselves flag the 0<γ≤1 regime as assumption-driven, and that is the most load-bearing weakness because it directly contradicts the advertised 'rigorous ... for any γ>0' claim. The numerical agreement with prior work is supportive but not a proof, and the validation data share authors with the paper. Therefore I recommend retaining the CONDITIONAL verdict: the upper bound is plausible and likely correct, but the proof as written is incomplete in the stated parameter range.","tokens_in":16455,"tokens_out":28802,"duration_ms":273987,"concrete_test":"For γ∈{0.25,0.5,0.75,1.0} compute the smallest eigenvalue of X_N = I − τ_up^{-γ} B_N for N up to 10^4 (using the compact operators B_i of Section III), and separately compute finite-N Gershgorin lower bounds d_m^{(N)} for all m<N at τ^γ=τ_up^γ. Check whether λ_min(X_N)>0 and d_m^{(N)}≥d_0^{(N)} for every m<N as N grows; also compare with the numerical solution of the truncated gap equation under two boundary conditions (θ_N=0 and θ_{N+1}=θ_N). If λ_min(X_N)<0 or the d_m^{(N)} inequality fails for any γ≤1, Eq. 44 is not a valid upper bound as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central new result is the closed-form upper bound Eq. 44, claimed valid for every γ>0. The proof route is: truncate H to N×N, apply a similarity transformation with O=diag(1/(n+p)), use Gershgorin on finite h^(N), then take N→∞ with p=1/2, relying on the appendix claim that the zeroth disc is lowest. Two load-bearing gaps remain. First, Section II.C explicitly states that for 0<γ≤1 the boundary condition lim_{m→∞} θ_m=0 is imposed, not proved, and that the untruncated-infinite-matrix calculation is an assumption supported only by agreement with numerics. Since the abstract and Section I advertise rigorous bounds for 'any γ>0', this is an admitted hole in exactly the regime where the infinite matrix and the gap equation are least controlled. Second, Section V passes to N→∞ with the remark that the Gershgorin proof extends under two conditions (existence of a discrete eigenvalue and bounded eigenvector entries), but does not supply the functional-analytic argument connecting finite-section eigenvalues of the unbounded H to the spectrum of the bounded compact-perturbation operator X. The appendix proof that τ_0^γ−τ_m^γ≥0 appears algebraically sound, but it is applied to infinite row sums whose convergence and domination of all finite-N discs are not fully demonstrated. If either gap cannot be closed, Eq. 44 could lie below the true τ_c for some γ, invalidating the claimed upper-bound status.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript derives analytical lower and upper bounds on the dimensionless transition temperature τ_c in the γ-model of quantum-critical superconductivity, building on the spin-chain representation of Migdal–Eliashberg theory. Lower bounds are obtained by applying Cauchy's interlacing theorem to finite truncations of the infinite Hessian H; the first four bounds are computed and agree with the lower bounds of Kiessling et al. [3]. The main new result is the closed-form upper bound τ_up^γ = Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1), obtained via a similarity transformation O = diag(1/(n+p)) with p=1/2 and Gershgorin circles, and claimed valid for all γ>0 and significantly tighter than existing bounds. The paper also discusses the justification of finite truncation through compact operators and Weyl's theorem.","tokens_in":16866,"tokens_out":23215,"duration_ms":184159,"significance":"If the proof can be completed, the upper bound is a substantial improvement over the existing bound in [3] and converges rapidly to numerical solutions. The lower-bound derivation via interlacing is clean and pedagogically appealing. The closed-form expression is explicit, parameter-free, and falsifiable by comparison with numerics. The compact-operator framework is inherited from [3] but is applied in a simpler and more direct way. The paper contains no fitted parameters; the p=1/2 choice is motivated analytically, though the claim of optimality is not fully demonstrated.","major_comments":[{"comment":"The paper advertises rigorous bounds for any γ>0, but Section II.C explicitly states that for 0<γ≤1 the boundary condition lim_{m→∞} θ_m=0 is imposed without proof, and that the N→∞ upper-bound calculation in this regime is an assumption supported only by numerical agreement. Since Eq. (44) and the central claim 'any γ>0' rely on this, the upper bound is not rigorously established for γ≤1. The authors must either prove the boundary condition and the discrete-spectrum assertion for γ≤1, or clearly restrict the rigorous claim to γ>1 and label the γ≤1 result as a conjecture.","section":"II.C, Abstract, Eq. (44)"},{"comment":"The Gershgorin argument is applied to finite N×N matrices h^{(N)}, and then the limit N→∞ is taken in the row sums. The paper merely states that the proof holds for N→∞ under two conditions, but does not supply the required spectral approximation: one must show that if every finite section H^{(N)} has only nonnegative eigenvalues at τ = τ_up, then the infinite Hessian H has no negative eigenvalue. This can be supplied using the compact-perturbation framework of Section III (finite-section convergence of eigenvalues of compact operators), but as written this is a gap that affects the status of Eq. (44) even for γ>1.","section":"V, Eq. (41)–(44)"}],"minor_comments":[{"comment":"Typos: 'Messiner' should be 'Meissner'; 'digonal' should be 'diagonal'; 'sectio,n' should be 'section'; spelling of Gershgorin/Gerschgorin is inconsistent.","section":"Throughout"},{"comment":"The caption states 'somewhere 1/2 < p < 1/3', which is an impossible inequality; presumably '1/6 < p < 1/3' is intended.","section":"Fig. 5 caption"},{"comment":"The conclusion claims 'Through an analytical optimization, we found the optimal value p=1/2.' The body and Appendix A only prove that p=1/2 makes the zeroth Gershgorin disc the lowest; they do not prove minimality of the resulting upper bound over p. This overstatement should be corrected.","section":"VI, Conclusion"},{"comment":"The closed forms for τ_{c,3} and τ_{c,4} are not given explicitly, only displayed in figures. If the paper advertises closed-form lower bounds, these expressions should be provided or the claim should be softened.","section":"IV.C"},{"comment":"The notation '1/2γ' in the definition of a is ambiguous; it should be written as 1/2^γ. Similarly, b should be 1/3^γ.","section":"Eq. (33)–(35)"},{"comment":"The upper bound is written as τ^γ < 2ζ(γ); since the Gershgorin argument allows equality, the bound should be τ^γ ≤ 2ζ(γ) (or the strict inequality justified).","section":"V.A, Eq. (38)"}],"recommendation":"major_revision","confidential_remarks":"The main issue is the mismatch between the abstract's 'rigorous for any γ>0' and the explicitly admitted assumption in Section II.C for 0<γ≤1. The result is likely correct—the numerical agreement is strong and the derivation for γ>1 is essentially sound—but the proof as written is incomplete. I do not see circularity or data-dependent fitting; the bound is standalone. The paper is within scope for cond-mat.supr-con, but the authors should either close the functional-analytic gaps or carefully restate the rigor claims."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this paper delivers something real. Eq. (44) is a closed-form upper bound on τc^γ that is a large improvement over the Kiessling et al. bound, and the derivation—similarity transformation plus Gershgorin circles with p=1/2—is clever and, on close reading, mostly sound. The lower-bound part is not new, since it reproduces the results in [3], but the interlacing derivation is a clean, accessible alternative. The appendix proof that the zeroth disc is the lowest for p=1/2 is actually there and, as far as I can tell, correct; the non-negativity argument is direct and applies to the infinite sums. So the core new result is more solid than the initial reviewer's low confidence would suggest.\n\nThe soft spots are real but manageable. First, the abstract says rigorous for any γ>0, but Section II.C admits that the boundary condition θ_m→0 is imposed, not proved, for 0<γ≤1, and that the infinite-matrix calculation in that regime is an assumption supported by numerical agreement. So the advertised rigor is not met there. The fix is simple: state the proven theorem for γ>1 and present the γ≤1 case as a physically motivated conjecture with numerical support. Second, the N→∞ step in the Gershgorin argument is brushed over. It is plausible given the compact-operator control in Sec. III, but a referee should ask for the finite-section convergence spelled out. That is a moderate fix, not a fundamental flaw. Third, the numerical comparison uses data from a paper with a common author. Not improper, but an independent benchmark would make the convergence claim stronger.\n\nBottom line: this is a useful analytical tool for the γ-model subfield, and the bound is far better than what was available. The authors are transparent about the γ≤1 limitation, which counts for something. If you work on quantum-critical pairing or Eliashberg-style bounds, this is worth a careful read. It deserves peer review, with a request to soften the abstract and tighten the limiting argument.","headline":"A genuinely new and much tighter upper bound on τc for the γ-model; the derivation is mostly sound, but the advertised 'any γ>0' rigor is ahead of the proof.","tokens_in":17325,"tokens_out":6401,"would_cite":true,"duration_ms":63724,"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":"In the γ-model of quantum-critical superconductivity, the transition temperature is bounded from above by a closed-form zeta-series expression, τ_c^γ ≤ Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1), claimed valid for every γ>0.","keywords":["quantum critical superconductivity","γ-model","Eliashberg theory","transition temperature bounds","Gershgorin circle theorem","Hessian eigenvalue analysis","Matsubara frequencies","non-Fermi liquid"],"falsifier":"Take a fixed γ in (0,1], form the N×N truncated Hessian at τ exactly equal to (Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1))^{1/γ}, and compute the smallest eigenvalue for increasingly large N. If for any N that eigenvalue is negative, or if the Gershgorin disk of some non-zeroth row dips below the claimed bound, then Eq. (44) is not a true upper bound; the claim is supported if the smallest eigenvalue stays non-negative and approaches zero from above as N grows.","tokens_in":16303,"feed_emoji":"❄️","tokens_out":5146,"duration_ms":55200,"temperature":0.7,"pith_summary":"The paper works with the γ-model, a minimal Eliashberg-style description of a metal at a quantum critical point in which the pairing interaction scales as 1/|Ω|^γ with no frequency cutoff. It asks how high the superconducting transition temperature can be as a function of the single parameter γ. The authors establish a closed-form upper bound: the dimensionless T_c^γ is no larger than Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1), claimed for all γ>0. They also reproduce lower bounds by checking when small truncations of the Hessian cease to be positive definite. A sympathetic reader cares because the calculation is parameter-free and because a sharp two-sided sandwich on T_c is one of the few rigorous statements available for strongly retarded pairing.","feed_headline":"Superconducting T_c capped by a simple zeta-series formula","feed_subtitle":"In the γ-model of pairing at a quantum critical point, T_c is sandwiched between tiny truncations and a rapidly converging zeta-series bound","key_machinery":"The carrying object is the linearized Hessian H of the spin-chain free energy functional, whose negative eigenvalue signals the pairing instability. Since H is unbounded, the paper maps it by a positive diagonal congruence to X̃ = I − τ^{-γ}B with B compact, so Weyl's essential-spectrum theorem confines sign-changing eigenvalues to the discrete spectrum and justifies finite truncation. Lower bounds come from Cauchy eigenvalue interlacing applied to nested principal submatrices; the upper bound comes from the Gershgorin circle theorem, which encloses all eigenvalues in disks centered at the diagonal entries, applied to a diagonal similarity transform of H. The similarity parameter is optimize","core_discovery":"The central claim is that the normal-state Hessian of the spin-chain free energy becomes positive definite at temperatures above a computable value, so no eigenvalue can cross zero and no superconducting instability can set in. That value is τ_up^γ = 1/2 Σ_{n=1}^∞ n^{-γ} 2n/((n+1/2)(n-1/2)) = Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1). The paper states the inequality τ_c^γ ≤ τ_up^γ holds for every γ>0 and shows numerically that this bound is far tighter than previous upper estimates, converging rapidly toward the values obtained by solving the gap equations.","pith_inferences":["One implication the authors leave implicit: if the convergence seen in their plots persists, the high-frequency tail of the pairing interaction contributes only a small correction to T_c, so a few low Matsubara modes effectively set the ordering temperature; this could be tested by comparing T_c from truncated frequency grids of different sizes.","The same diagonal-similarity Gershgorin construction is likely portable to related Eliashberg problems with dispersive or Einstein phonons, where the corresponding upper bounds are looser; a numerical check would be to apply the p=1/2 trick to those kernels and compare with large-truncation eigenvalues.","Because the paper imposes the boundary condition lim_{m→∞} θ_m = 0 for 0<γ≤1 without proof, the 'any γ>0' claim should be read as conditional until a direct proof appears or until a large-truncation numerical counterexample at tiny γ is ruled out."],"forward_implications":["For any γ>0, the dimensionless transition temperature is sandwiched between the finite-truncation lower bounds and the closed-form upper bound, so T_c can be located without solving the full nonlinear Eliashberg equations.","Because the γ-model has only one energy scale g, this yields a universal function f(γ) such that k_B T_c = g f(γ), with the allowed interval for f(γ) now narrow enough to be practically predictive.","The upper bound remains finite and convergent even for 0<γ≤1, a regime where the bare Matsubara sums in the gap equation diverge and where previous closed-form estimates were loose or unavailable.","The determinant-zero conditions for the 1×1 through 4×4 truncations independently reproduce earlier variational lower bounds, giving a more elementary route to the same result."],"fun_headline_variants":["Zeta-series pin T_c in γ-model","Tight T_c ceiling from a zeta sum","Rigorous T_c bound via ζ-series","Simple zeta formula sets T_c upper bound"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that for 0<γ≤1 the boundary condition lim_{m→∞} θ_m = 0 can be imposed without proof, and that for p=1/2 the zeroth Gershgorin disk remains the lowest in the infinite-N limit; if either fails, τ_up^γ may sit below the true τ_c instead of above it.","fun_headline_variants_meta":{"raw":{"variants":["Zeta-series pin T_c in γ-model","Tight T_c ceiling from a zeta sum","Rigorous T_c bound via ζ-series","Simple zeta formula sets T_c upper bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000992,"raw_usage":{"total_tokens":4066,"prompt_tokens":797,"completion_tokens":3269,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":541,"completion_tokens_details":{"reasoning_tokens":3209}},"tokens_in":541,"tokens_out":3269,"duration_ms":22557,"temperature":1.0,"reasoning_tokens":3209,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T14:30:13.048920+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a fixed γ in (0,1], form the N×N truncated Hessian at τ exactly equal to (Σ_{n=0}^∞ 4^{-n} ζ(γ+2n+1))^{1/γ}, and compute the smallest eigenvalue for increasingly large N. If for any N that eigenvalue is negative, or if the Gershgorin disk of some non-zeroth row dips below the claimed bound, then Eq. (44) is not a true upper bound; the claim is supported if the smallest eigenvalue stays non-negative and approaches zero from above as N grows.","supporting_citations":[],"review_version":1}