{"id":"f5311d6e-a837-4a39-8a57-147800f90f65","arxiv_id":"2505.14498","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Periodic Jacobi operators on the half-line satisfy t^{-1/2} weighted dispersive decay in full generality and t^{-1/3} or t^{-1/(q+1)} global decay under explicit spectral conditions.","lead":"Periodic Jacobi operators on the discrete half-line are shown to have dispersive decay: solutions spread with at least t^{-1/2} decay in a weighted norm, with global t^{-1/3} or t^{-1/(q+1)} rates under stated conditions. The proof uses an explicit orthogonal-polynomial expansion of the propagator instead of a Fourier transform, which is unavailable on the half-line.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The t^{-1/3} proof in §5.2 has a near-resonance uniformity gap, but it is repairable by applying Van der Corput with s=3 directly; Theorem 3.1 appears sound.","rationale":"The reader's weakest-assumption hunt focused on the imported spectral fact that t_{2,1} is nonvanishing and sign-constant on each band, and on the generic nondegeneracy k_j''=0 => k_j''' \\neq 0. I do not find the t_{2,1} concern to be load-bearing: the cited theorem and the integrability of the spectral density exclude interior zeros, and the paper's Q(y) in (4.9) is therefore smooth on each band. The genuinely load-bearing issue, matching the reader's secondary concern, is the proof of part (1) of Theorem 3.2: the argument as written bounds F^{(s)}_0 only at a selected sequence of times and claims a uniform first-derivative lower bound for all other times, which fails near resonance. This is a real gap in the written proof, but it is not fatal to the theorem, because the support of X_j was engineered to avoid T_3, so |k_j'''| is bounded below and Van der Corput with s=3 gives the t^{-1/3} decay for every t and every integer frequency \\ell. Thus the central local estimate Theorem 3.1 stands, and the global theorem is correct modulo a local proof repair. The verdict CONDITIONAL is therefore appropriate, unchanged from the reader's assessment, since the paper should be revised to replace the sequence argument in §5.2 with the direct third-derivative bound.","tokens_in":17135,"tokens_out":22455,"duration_ms":232652,"concrete_test":"Re-derive the bound for F^{(s)}_0 in §5.2 without the sequence t_i: on the support of X_j, verify that |k_j'''(\\varphi)| \\geq c > 0 follows from the definition of \\eta in (5.12), then apply Lemma 5.1 with s=3 to the integral in (5.16) for every t and every integer \\ell, checking that the constant is independent of n, m, and \\ell. If the bound holds uniformly, the t^{-1/3} claim of Theorem 3.2(1) is established and the manuscript only needs a revised passage; if the constant instead depends on the quality of rational approximations of k_j'(\\varphi_p), a genuine near-resonance condition is required.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing issue is not in the local theorem but in the proof of Theorem 3.2(1). After the identity (5.15), the integral for F^{(s)}_0 becomes a sum of terms of the form (5.16), with phase \\lambda(\\varphi) = -t k_j(\\varphi) + \\ell \\varphi and \\ell \\in \\mathbb{Z}. The manuscript selects a sequence t_i, \\ell_i for which \\partial_\\varphi\\lambda and \\partial_\\varphi^2\\lambda vanish at a point \\varphi_p with k_j''(\\varphi_p)=0, applies Lemma 5.1 with s=3 only on this sequence, and then claims that for every other time a lower bound on |\\partial_\\varphi\\lambda| makes Lemma 5.2 yield a t^{-1} rate. This claim is not uniform: for any neighborhood of a resonant time, there are integers \\ell making |t k_j'(\\varphi) - \\ell| arbitrarily small, so no positive lower bound on the first derivative exists independently of t and \\ell. However, the gap is repairable: the cutoff X_j in (5.12) was chosen so that its support contains no points of T_3 and no endpoints, hence |k_j'''| \\geq c > 0 there. Since \\lambda''' = -t k_j''', Lemma 5.1 with s=3 applies directly to (5.16) for all t and all \\ell, yielding the t^{-1/3} bound uniformly. Thus the theorem is correct, but the written proof contains a false uniformity assertion that should be replaced by this direct VdC argument. For Theorem 3.1, no comparable obstruction appears: stationary contributions come only from gapped band edges, away from which |k_j'| has a uniform positive lower bound, and the imported nonvanishing of t_{2,1} on bands is standard and consistent with integrability of the spectral density.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies dispersive decay for e^{-itJ}P_c where J is a periodic Jacobi operator on ℓ²(N). Using the explicit eigenfunction expansion of J in terms of transfer matrices and the discriminant Δ, the authors reduce the propagator to finite sums of oscillatory integrals with phase -t k_j(φ)+ℓφ. Theorem 3.1 claims a weighted ℓ¹₁→ℓ^∞_{-1} bound t^{-1/2} for every period q. Theorem 3.2 claims a global ℓ¹→ℓ^∞ bound t^{-1/3} under the nondegeneracy condition 'if k_j''=0 then k_j'''≠0', and t^{-1/(q+1)} for even q when the spectrum has exactly q disjoint bands. The paper is self-contained after importing standard spectral facts on periodic Jacobi operators.","tokens_in":17399,"tokens_out":24144,"duration_ms":232756,"significance":"If valid, Theorem 3.1 appears to be the first dispersive estimate of this type for periodic Jacobi operators on the half-line and gives a uniform t^{-1/2} local decay independent of the band structure; the global results connect to known rates for periodic Schrödinger operators on Z and to edge-mode asymptotics. The proof strategy is transparent and parameter-free: it reduces the problem to controlled oscillatory integrals and explicitly identifies the roles of gapped versus ungapped band edges. The paper also honestly states open questions, e.g., genericity of the nondegeneracy condition and the correct rate when it fails, and it cites parallel work [14]. However, as detailed below, the written derivations of the simplified amplitude and of the uniformity in Theorem 3.2(1) need correction.","major_comments":[{"comment":"The change of variables leading to the simplified propagator is not correct as written. Since √(4−Δ²)=2|sinΘ| with Θ∈[−π,0], the density (2.10) is proportional to (−sinΘ)/t_{2,1} dx, not to 1/(t_{2,1} sinΘ) dx as printed in (4.7). Moreover, from (4.10), k′_j(φ)=−2 sinφ/Δ′_j(k_j(φ)), so dx=−2 sinφ/Δ′_j(k_j(φ)) dφ, not Δ′/(−2 sinφ)dφ as stated before (4.8). With these corrections one obtains an amplitude proportional to L/(t_{2,1}Δ′) rather than Q=Δ′L/t_{2,1} in (4.8)–(4.9). Because Q is used in every subsequent Van der Corput estimate, this algebraic discrepancy must be resolved; as it stands the derivation of (4.8) does not follow from (4.6).","section":"Section 4.1, Eqs. (4.7)–(4.9)"},{"comment":"The proof of Theorem 3.2(1) contains a uniformity gap. After (5.15), the phase in (5.16) is λ(φ)=−t k_j(φ)+ℓφ. The text states that for all t≠t_i there is a lower bound on |∂_φ λ| and hence Lemma 5.2 gives a t^{-1} rate; this is false uniformly in ℓ and t, because for any t one can choose ℓ∈Z with t k′_j(φ_p)−ℓ arbitrarily small. The gap is repairable: the cutoff η in (5.12) keeps the support of X_j away from T_3, so |k_j'''|≥c>0 there, and since λ'''=−t k_j''', Lemma 5.1 with s=3 applies directly to each term in (5.16) for every t and every ℓ. The theorem is plausible, but the proof as written should be replaced by this direct application.","section":"Section 5.2, Eqs. (5.15)–(5.17)"},{"comment":"The final paragraph of the proof of Theorem 3.2(2) is too compressed to verify the claimed t^{-1/(q+1)} rate. It says that a point where only a nonzero lower bound on k^{(q)} is available yields t^{-1/(q+1)}, but Lemma 5.1 with s=q gives t^{-1/q}; also the partition of [−π,0] according to which derivative is bounded below is not constructed, and the uniformity of the constants and cutoffs is not shown. Please expand this proof or clarify the exponent.","section":"Section 5.2, proof of (3.2)"}],"minor_comments":[{"comment":"The phrase 'assume without loss of generality that t1,2 > 0' should refer to t_{2,1}, the entry that appears in the density formula (2.10); this is a typo in the index.","section":"Section 4.1, after Eq. (4.6)"},{"comment":"The definition of η uses a minimum over T_2, but the case T_2=∅ is not handled; if k_j'' has no zeros then F^{(s)}_0 is identically zero and the argument is trivial, but this case should be stated explicitly.","section":"Section 5.2, Eq. (5.12)"},{"comment":"The line 'k(ℓ)_j(φ0)=0 for for some 2≤ℓ≤q' should read 'for every 2≤ℓ≤q', and the duplicated 'for' should be removed.","section":"Lemma 4.7, proof"},{"comment":"The phrase 'J has only continuous spectrum and pure point spectrum' is imprecise; it should say 'absolutely continuous spectrum and pure point spectrum', since the continuous spectrum is shown to be purely absolutely continuous.","section":"Theorem 2.1(2)"},{"comment":"The claimed 'more elaborate and concrete computation of the propagator' for the SSH model is not identified; a pointer to the relevant equations or a short derivation would help the reader verify the t^{-1/3} claim.","section":"Remark 3.3"}],"recommendation":"major_revision","confidential_remarks":"The algebraic discrepancy in (4.7)–(4.9) is probably a typo and the uniformity gap in §5.2 is repairable by the direct Van der Corput argument described in the report; I do not see a reason to doubt the truth of the main results. Nevertheless, because the oscillatory integral representation is the foundation of every proof, the manuscript needs a corrected derivation and a rewritten proof of Theorem 3.2(1) before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper proves the first dispersive decay estimates for periodic Jacobi operators on the half-line. The main result, Theorem 3.1, is a local weighted ℓ∞_{-1} bound of t^{-1/2} for all periods, and it is new. The proof via orthogonal polynomial spectral representation plus Van der Corput estimates is careful and, as far as I can check, sound. The global results in Theorem 3.2 are plausible, but the written proof of part (1) has a real gap.\n\nSpecifically, after decomposing the oscillatory integral, the authors pick a sequence of times where both the first and second derivatives of the phase vanish at a point φ_p with k''(φ_p)=0, apply Van der Corput with s=3 only on that sequence, and then claim that all other times have a uniform lower bound on the first derivative, giving a faster t^{-1} decay. That claim is false uniformly in the integer ℓ = ±(n±m), because for any t there are ℓ making t k'(φ) ≈ ℓ, so no positive lower bound exists that is independent of t and ℓ.\n\nHowever, the gap is repairable. The cutoff X_j in (5.12) was chosen so that its support contains no zeros of k''' and no endpoints, hence |k'''| ≥ c > 0 there. Since the third derivative of the phase λ(φ) = -t k_j(φ) + ℓ φ is -t k''', Lemma 5.1 with s=3 applies directly to the integral (5.16) for all t and all ℓ, yielding the desired t^{-1/3} bound uniformly. So Theorem 3.2(1) is correct, but the proof as written needs this fix.\n\nThe even-q result (Theorem 3.2(2)) relies on Lemma 4.7, which seems sound, though the final partitioning argument is compressed. The imported spectral facts, especially the nonvanishing of t_{2,1} on bands, are standard and appear reliable. The paper is honest about the nondegeneracy condition being unproven generically, and there is no circularity or fitting. The citation pattern is appropriate, including the parallel work of Damanik, Fillman, and Young on Z.\n\nWho is this for? Spectral theorists and mathematical physicists working on edge transport and Floquet materials. The local theorem is the cleanest contribution and should stand. I recommend sending this to peer review; a referee should ask the authors to fix the uniformity argument in Section 5.2, which is a short repair, not a change in the result.","headline":"The local t^{-1/2} decay is solid and new; the global t^{-1/3} proof has a repairable uniformity gap.","tokens_in":18038,"tokens_out":3328,"would_cite":true,"duration_ms":31252,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":null,"created_at":"2026-08-07T15:34:29.440279+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":null,"supporting_citations":[],"review_version":1}