{"id":"cec6aae6-d6ed-4612-b9f5-dd63ded42ed1","arxiv_id":"2608.11850","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A perturbed cubic Alltop state yields minimal Weyl-Heisenberg measurements whose spectrum is confined to a narrowing interval and whose weakest direction reaches the symmetric-informationally-complete benchmark asymptotically.","lead":"This paper constructs explicit quantum measurements whose statistical stability can be made almost as good as the ideal symmetric measurement, without needing that ideal to exist. It proves that in many dimensions the weakest operator direction is uniformly well resolved and the whole spectrum becomes flat.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified; the load-bearing Gauss-sum flatness is standard and exactly satisfied, and Theorem 11's interval, exact floor, and optimality squeeze hold under re-derivation.","rationale":"I stress-tested the central claim by re-deriving the three links on which Theorem 11 depends. First, the reader's identified crux is correct: the finite-field quadratic Gauss sum (F7), used in Lemma 9, is the only external input that makes the Alltop ambiguity profile exactly flat. I verified that the identity is standard, correctly stated, and correctly applied in both cyclic and extension fields, so the flatness is exact rather than approximate. Second, I re-derived Lemma 10's repair bounds with the paper's exact displacement conventions, including the seemingly delicate cross terms; they match the printed formulas (80) and (F8), and the bounds depend only on moduli, making the exact floor λ(φ_q)=L_q and the interval [L_q,U_q] robust. Third, the balance equation, the monotonicity argument for L_q ≥ L_5, and the squeeze L_q ≤ Λ*_q ≤ q/(q+1) are elementary and do not assume SIC existence. I also spot-checked q=5 numerically: all off-axis eigenvalues fall inside [L_5,U_5], the repaired axis attains L_5, and the full spectrum is consistent with the Moyal sum. The exclusion of characteristics 2 and 3 is honest and correctly motivated by the degenerate quadratic coefficient; the binary case is handled separately and correctly, and characteristic three is disclosed as open. I therefore found no load-bearing concern and no internal inconsistency. The reader's ACCEPT verdict, with moderate confidence and medium correctness risk, remains appropriate; the only assumption worth an independent check is the Gauss-sum flatness in extension fields, which my proposed concrete test targets.","tokens_in":22054,"tokens_out":44087,"duration_ms":384774,"concrete_test":"Invert the dependency test: over F_25 (extension field, p=5), enumerate all 624 nonzero (a,b) and compute |χ_A(a,b)|² = |q^{-1}Σ_{x∈F_25} ψ(-3ax²+(b-3a²)x)|² by direct summation, verifying exact flatness (0 on a=0, b≠0; 1/25 for all a≠0). Separately, for q=5 and q=7, enumerate all q²-1 nonidentity projector-Gram eigenvalues of φ_q symbolically and confirm min = L_q and max ≤ U_q. Both checks would immediately expose any hidden failure of the Gauss-sum input or of Lemma 10's bound; the paper's own F25/F49 numerics purport to pass, but an independent enumeration settles it.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No load-bearing objection found. The central claim (Theorem 11) rests on three links: (i) Lemma 9's exact flatness |χ_A(a,b)| = q^{-1/2} for all a≠0, from the finite-field quadratic Gauss sum (F7); (ii) Lemma 10's triangle-inequality repair bounds; (iii) the balance/monotonicity calculus and the Λ* squeeze. I re-derived each link with the paper's displacement convention D_{a,b}=X_aZ_b, Z_b|x⟩=ψ(bx)|x⟩, including the complete off-axis formula (F8) and the repaired-axis value (83): the ψ(-a³) prefactor and the 3a³/4 → -a³/4 phase absorption are consistent, and the cross terms tψ(-a³)[1+ψ(-ab)] are exactly what ⟨0|D|A⟩ and ⟨A|D|0⟩ produce. Equation (80) itself checks against direct enumeration for q=5, e.g., χ(1,0) = ψ(-1)/5·Σψ(-3x²-3x) matches ⟨A|X|A⟩ numerically. Because Lemma 10 uses only moduli, the off-axis lower bound (1-2t)²/N² is valid regardless of phases, so λ(φ_q)=L_q is exactly attained on the (0,b≠0) axis and the interval [L_q,U_q] follows. The squeeze L_q ≤ Λ*_q ≤ q/(q+1) is elementary and yields λ/Λ*→1 without SIC assumptions. Spot computation for q=5 (all 24 nonidentity eigenvalues bounded, axis equal to 0.19786) is consistent. The dependency the reader flags is real but satisfied: in characteristics 2 and 3 the quadratic coefficient -3a vanishes or 4A is non-invertible, so flatness genuinely breaks; the paper excludes these, treats p=2 separately, and candidly leaves p=3 open. I found no internal inconsistency and no overreach in the claimed operational corollaries.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the spectral stability of minimal rank-one Weyl–Heisenberg (WH) measurements, quantified by the smallest nonidentity eigenvalue λ of the projector-Gram matrix. It proves several explicit constructions: a Haar-generic instability result (E[λ^{-1}]=∞), a cyclic family with polynomial floors Θ(d^{-3}) in odd and Θ(d^{-5}) in even dimensions, a characteristic-two finite-field family with uniform λ≥4/9, and the main result: for every prime power q=p^r with p≥5, a balanced one-coordinate perturbation of the cubic Alltop state yields a fiducial whose nonidentity spectrum lies in [L_q,U_q] with L_q≥L_5≈0.1979, U_q/L_q→1, η(φ_q)→1, and λ(φ_q)/Λ*_q→1, all without assuming SIC existence. Exact finite-sample Hilbert–Schmidt MSE, canonical-shadow bounds, and Fisher-efficiency relations are derived from the spectrum.","tokens_in":22468,"tokens_out":16800,"duration_ms":151889,"significance":"The main theorem is a genuinely significant contribution: it gives an explicit unconditional family of minimal WH measurements whose spectral floor approaches the SIC max–min benchmark in finite-field dimensions. The proofs are self-contained, the load-bearing Gauss-sum flatness is standard and exactly satisfied, and no numerical optimization or SIC-existence assumption enters. The characteristic-two and cyclic results are also exact and supply useful intermediate benchmarks. The operational corollaries (canonical MSE, Fisher efficiency, shadow bounds) follow directly from the spectral information and are clearly scoped. If the results stand as written, this substantially advances the quantitative design of informationally complete measurements.","major_comments":[],"minor_comments":[{"comment":"The identity N_q = 1 + a_q/s in Eq. (94) is stated without derivation; it follows from the balance equation s t^2 + 4t = 1 and deserves a one-line explanation.","section":"III C (Eq. 94)"},{"comment":"The DOI \"10.1103/wkj6-l7bf\" in Ref. [8] appears to be a placeholder and should be corrected before publication.","section":"References"},{"comment":"In the sharpness argument for the even-dimensional cyclic family, the case d=4 is not explicitly handled; the sentence saying the bound follows trivially from λ≤d/(d+1) is correct, but a direct check for d=4 would improve readability.","section":"Appendix D"},{"comment":"The caption of Fig. 1(c) states \"Exact balanced-spike formula\"; I suggest clarifying that the plotted curve is the proven attained floor L_q, not a numerical estimate, to match the theorem statements.","section":"V (Fig. 1 captions)"}],"recommendation":"minor_revision","confidential_remarks":"No concerns about citation patterns or scope. The manuscript fits the journal well. The only issue I would flag separately is the placeholder DOI in Ref. [8], which should be fixed. I found no load-bearing technical errors; the requested changes are local and presentation-only."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Bottom line: this paper earns a serious referee. The headline result—an explicit balanced-Alltop fiducial in every finite-field dimension of characteristic p≥5 whose entire nonidentity projector-Gram spectrum lies in a shrinking interval [Lq,Uq], with a uniform floor at L5≈0.1979 and λ(φ_q)/Λ*_q→1—survives re-derivation. The stress-test note is right: the Gauss-sum flatness is exactly satisfied, the repair lemma uses only moduli, and the squeeze to the global max–min optimum is elementary once you have the interval.\n\nWhat is genuinely new is the balanced one-coordinate perturbation: t_q from the analytic crossing of the repaired-axis and bulk distortion, the closed interval, the strict monotonicity in q, and the unconditional asymptotic optimality. The cyclic family (Θ(d^{-3}) odd, Θ(d^{-5}) even) and the characteristic-two family with uniform floor 4/9 are useful additions. The paper also does something rare: it gives closed-form spectra, not numerically fitted constants. I checked the off-axis formula (F8) and the q=5 values; they are consistent. The citations look right, and the spectral dictionary credit to Goldberger et al. is explicit.\n\nSoft spots, in proportion. The proofs are dense and I did not machine-check them; the finite-field character calculations have enough moving signs that a formal verification would be reassuring. Characteristic three is honestly left open. The characteristic-two family is uniformly stable but not asymptotically flat, so the \"approaching SIC benchmark\" claim is really about p≥5. The operational corollaries are scoped to the maximally mixed state (exact MSE) and to worst-direction uniform bounds; that is fine but easy to over-read. The convergence rate q^{-1/4} is slow, but that is not a flaw.\n\nThe paper is for quantum information people working on tomography, frames, or approximate SICs. I would bring it to a reading group and would cite it. It deserves peer review, not desk rejection.","headline":"A solid, explicit construction with closed-form spectra that gives an unconditional near-SIC stability floor in finite-field dimensions; worth refereeing.","tokens_in":22994,"tokens_out":2254,"would_cite":true,"duration_ms":24168,"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":"Explicit finite-field measurements are shown to approach SIC-level stability without assuming SICs exist.","keywords":["Weyl–Heisenberg measurements","informationally complete POVMs","spectral stability","SIC benchmark","Alltop states","finite-field quadratic Gauss sums","quantum tomography","classical shadows"],"falsifier":"Construct the full nonidentity projector-Gram spectrum of $|\\phi_q\\rangle$ by exact arithmetic for $q=5$, $q=7$, and $q=25$, either by direct diagonalization or from the closed formula in the paper's Appendix F. Theorem 11 predicts that the minimum eigenvalue equals $L_q$, that every off-axis eigenvalue lies in $[L_q,U_q]$, and that the multiplicity of $L_q$ is at least $q-1$; a single off-axis eigenvalue below $L_q$, or a computed minimum different from $L_q$, would refute the exact-floor claim.","tokens_in":21840,"feed_emoji":"⚛️","tokens_out":14089,"duration_ms":133754,"temperature":0.7,"pith_summary":"This paper asks how well-conditioned a minimal quantum measurement can be made without assuming that an exact symmetric informationally complete (SIC) measurement exists. For every prime-power dimension $q=p^r$ with $p\\ge 5$, it constructs an explicit Weyl–Heisenberg fiducial—a balanced one-coordinate perturbation of a cubic Alltop state—whose nonidentity projector-Gram spectrum lies in a short interval $[L_q,U_q]$ with $L_q\\ge 0.1979$ and $U_q/L_q\\to 1$. The SIC-normalized minimum $\\eta(\\phi_q)$ tends to $1$, and the attained floor $\\lambda(\\phi_q)$ tends to the global finite-field Weyl–Heisenberg max–min optimum $\\Lambda_q^\\star$, all without assuming SIC existence. If the paper is right, this gives explicit, dimension-uniformly stable minimal informationally complete measurements whose full error spectrum asymptotically matches the SIC benchmark, including the exact finite-sample tomographic error at the maximally mixed state.","feed_headline":"Balanced Alltop states nearly hit the SIC stability benchmark","feed_subtitle":"One repaired cubic state gives asymptotically flat spectra, approaching the SIC benchmark.","key_machinery":"The load-bearing object is the balanced-Alltop fiducial $|\\phi_q\\rangle=(|A_q\\rangle+t_q|0\\rangle)/\\sqrt{N_{q,t_q}}$, where $|A_q\\rangle=q^{-1/2}\\sum_{x\\in\\mathbb{F}_q}\\psi(x^3)|x\\rangle$ is the cubic Alltop state. The Alltop ambiguity profile $|\\chi_{A_q}(a,b)|$ is exactly $q^{-1/2}$ for every $a\\neq 0$ and zero on the axis $a=0$, $b\\neq 0$; the finite-field quadratic Gauss-sum identity is what makes this profile exact. The one-coordinate spike repairs the zero axis, and the balance equation $\\sqrt{q}\\,t^2+4t=1$ (equivalently $t_q\\sim q^{-1/4}$) equalizes the two competing scales—linear bulk distortion versus quadratic repaired-axis amplitude. The spectral theorem for Weyl–Heisenberg projector Gram matrices, $G_\\phi|f_{m,n}\\rangle = q|\\chi_\\phi(-n,m)|^2|f_{m,n}\\rangle$, then converts the repaired ambiguity profile into the interval $[L_q,U_q]$ and identifies the exact floor $\\lambda(\\phi_q)=L_q$.","core_discovery":"The central claim is that spectral flatness, not just completeness, can be achieved by an explicit construction. The unperturbed cubic Alltop state over $\\mathbb{F}_q$ has ambiguity profile exactly flat at magnitude $q^{-1/2}$ except for one zero axis, so its Weyl–Heisenberg orbit is incomplete; adding a spike $t|0\\rangle$ repairs that axis. The paper shows that choosing $t_q\\sim q^{-1/4}$ balances the repaired-axis amplitude against the distortion of the flat bulk, producing a fiducial $\\phi_q$ whose smallest nonidentity projector-Gram eigenvalue equals the certified bound $L_q$, with the whole nonidentity spectrum confined to $[L_q,U_q]$. Consequently $\\eta(\\phi_q)=(q+1)\\lambda(\\phi_q)/q\\to 1$ and $\\lambda(\\phi_q)/\\Lambda_q^\\star\\to 1$, where $\\Lambda_q^\\star$ is the maximum over all normalized finite-field Weyl–Heisenberg fiducials, and this asymptotic optimality is unconditional on SIC existence. The paper also organizes explicit constructions into a hierarchy: polynomial floors $\\Theta(d^{-3})$ and $\\Theta(d^{-5})$ in cyclic dimensions, a uniform floor $\\lambda\\ge 4/9$ in characteristic two, and the near-SIC interval in characteristic at least five.","pith_inferences":["Inference: The zero-axis repair in Lemma 10 is stated abstractly for any flat-profile phase state, so the same one-coordinate perturbation could be tried on other Alltop-type profiles, with a different balance scale; the paper only suggests this direction.","Inference: Because characteristic 3 is excluded only by the degeneracy of the quadratic Gauss sum, a separate construction for $q=3^r$ is the natural missing piece of the prime-power picture; the cubic mechanism cannot be transplanted there as proven.","Inference: The limit $\\lambda(\\phi_q)/\\Lambda_q^\\star\\to 1$ suggests that in these dimensions the finite-field max–min problem is nearly solved by an explicit non-symmetric construction; whether the finite-$q$ optimum is attained only at an exact SIC remains open.","Inference: The near-isotropy of the full spectrum suggests that error bounds away from the maximally mixed state might also approach SIC-level performance, but the paper does not establish state-dependent bounds beyond $\\rho=I/q$."],"forward_implications":["For every prime-power dimension $q=p^r$ with $p\\ge 5$, there is an explicit minimal $q^2$-outcome informationally complete POVM whose nonidentity spectrum is asymptotically isotropic: $U_q/L_q\\to 1$ and the traceless condition number tends to $1$.","At the maximally mixed state, canonical linear inversion has finite-sample Hilbert–Schmidt mean-squared error ratio $R(\\phi_q)$ between $d/((d+1)U_q)$ and $d/((d+1)L_q)$, and both endpoints tend to $1$, so the explicit construction asymptotically matches the SIC inversion error.","The worst-direction local Fisher information at $I/d$, normalized by the SIC value, is $\\eta(\\phi_q)\\to 1$, meaning no traceless parameter direction is asymptotically harder to estimate than in a SIC.","Canonical classical-shadow estimators of $K$ observables achieve target accuracy with sample complexity proportional to $(d+1)/\\eta(\\phi_q)$, and the universal overhead $\\eta^{-1}$ is bounded by a dimension-independent constant for this family.","The characteristic-two family covers every multi-qubit dimension with uniform floor $\\lambda\\ge 4/9$, and at $q=2$ and $q=8$ it reaches the finite-field SIC endpoint exactly."],"supporting_citations":[{"why":"diagonalizes the rank-one projector Gramian by the two-dimensional Fourier transform, giving the spectrum $q|\\chi_\\phi(u)|^2$ used throughout.","marker":"[1]"},{"why":"defines SIC-POVMs and supplies the symmetric endpoint against which the max–min benchmark is measured.","marker":"[9]"},{"why":"provides the geometric all-dimensional BIC construction whose exponentially decaying spectral floor motivates the uniform-stability problem.","marker":"[14]"},{"why":"introduces the cubic Alltop sequences on which the base state $|A_q\\rangle$ is built.","marker":"[19]"},{"why":"supplies the finite-field quadratic Gauss-sum identity that makes the unperturbed Alltop ambiguity profile exactly flat.","marker":"[22]"}],"fun_headline_variants":["Repaired Alltop states yield near-SIC spectral flatness","Explicit minimal WH measurements approach SIC benchmark","Balanced Alltop states get near-constant spectra","Cubic Alltop repair approaches the SIC stability bound"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"All the main claims depend on an exact identity for quadratic exponential sums over finite fields; if that identity degenerates or holds only approximately, the flatness the construction repairs—and with it the certified interval $[L_q,U_q]$ and the equality $\\lambda(\\phi_q)=L_q$—would not follow.","fun_headline_variants_meta":{"raw":{"variants":["Repaired Alltop states yield near-SIC spectral flatness","Explicit minimal WH measurements approach SIC benchmark","Balanced Alltop states get near-constant spectra","Cubic Alltop repair approaches the SIC stability bound"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000924,"raw_usage":{"total_tokens":4063,"prompt_tokens":1148,"completion_tokens":2915,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":764,"completion_tokens_details":{"reasoning_tokens":2850}},"tokens_in":764,"tokens_out":2915,"duration_ms":23342,"temperature":1.0,"reasoning_tokens":2850,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:25:30.916273+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct the full nonidentity projector-Gram spectrum of $|\\phi_q\\rangle$ by exact arithmetic for $q=5$, $q=7$, and $q=25$, either by direct diagonalization or from the closed formula in the paper's Appendix F. Theorem 11 predicts that the minimum eigenvalue equals $L_q$, that every off-axis eigenvalue lies in $[L_q,U_q]$, and that the multiplicity of $L_q$ is at least $q-1$; a single off-axis eigenvalue below $L_q$, or a computed minimum different from $L_q$, would refute the exact-floor claim.","supporting_citations":[{"cited_title":"Substitution of a character gives Gϕfm,n =  ∑ a,b gϕ(a,b)ωma+nb  fm,n.(A1) The d2 characters are orthonormal and complete, while Eq","cited_arxiv_id":null,"evidence_quote":"diagonalizes the rank-one projector Gramian by the two-dimensional Fourier transform, giving the spectrum $q|\\chi_\\phi(u)|^2$ used throughout."},{"cited_title":"Acharya, S","cited_arxiv_id":null,"evidence_quote":"defines SIC-POVMs and supplies the symmetric endpoint against which the max–min benchmark is measured."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"introduces the cubic Alltop sequences on which the base state $|A_q\\rangle$ is built."},{"cited_title":"Singal, F","cited_arxiv_id":null,"evidence_quote":"supplies the finite-field quadratic Gauss-sum identity that makes the unperturbed Alltop ambiguity profile exactly flat."}],"review_version":1}