{"id":"6f9bfc55-6a92-4821-9354-10fbc8ab8f04","arxiv_id":"2607.02778","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Modified-Bessel kernels Is(x) for real s≥0 are STP∞, and totally positive Pólya-frequency sequences are dense in the product topology on two-sided PF sequences.","lead":"The paper proves the modified-Bessel kernel with real orders is strictly totally positive of infinite order, and that totally positive Pólya-frequency sequences are dense among all such sequences under pointwise limits. These settle two open questions in total-positivity theory via a spectral Darboux induction and discrete Gaussian Toeplitz smoothing.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified","rationale":"The manuscript answers two published open questions with short, self-contained classical arguments. The only step that is family-specific—the survival of ordered left-endpoint asymptotics through Darboux transforms—is written out inductively for the modified-Bessel solutions and does not rely on an unproved general principle. Question 5.1 correctly isolates the open generalization and does not affect Theorems 1.1–1.2. The product-topology density statement is likewise complete within the topology it claims. No circularity, missing estimate, or unstated hypothesis appears. The reader’s ACCEPT / HIGH-confidence verdict is therefore left unchanged.","tokens_in":10312,"tokens_out":433,"duration_ms":4882,"concrete_test":"For a concrete low-order check, evaluate the 2\times2 minor det[I0(1) I1(1); I0(2) I1(2)] numerically from the power series or a standard library; the value is positive (≈0.21). The same verification for a non-integer pair (e.g., s=0.5,1.5) confirms the real-order claim at the smallest nontrivial size.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The reader's weakest-assumption note correctly flags the inductive propagation of left-endpoint asymptotics under Darboux stages as the only non-generic step in the Bessel argument. That step is, however, carried out explicitly for the modified-Bessel family (power-series expansion of Is, termwise differentiation, and the inductive display after “Fix 0≤s1<⋯<sm”). The resulting leading coefficients remain positive products of (sj-sℓ) factors, so the hypotheses of Proposition 3.1 are verified rather than assumed. The Toeplitz-smoothing argument is likewise classical Cauchy–Binet plus the Aissen–Schoenberg–Whitney–Edrei representation and contains no hidden gap. Consequently neither central claim rests on an insecure premise.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper isolates two strictification mechanisms for total positivity and applies each to resolve an open question. First, a spectral Darboux induction (Proposition 3.1) converts positivity plus ordered left-endpoint asymptotics of a one-dimensional Sturm family into positive Wronskians and hence STP∞; applied to fs(y)=Is(e^y), this yields Theorem 1.1: the modified-Bessel kernel K(x,s)=Is(x) is STP∞ on (0,∞)×[0,∞), answering the real-order question of Buchstaber–Glutsyuk. Second, discrete convolution with the strictly totally positive Toeplitz kernel q^{n^{2}} (0<q<1), together with a Gaussian tilt for geometric sequences, shows that every two-sided Pólya-frequency sequence is a pointwise limit of totally positive ones (Theorem 1.2), answering Question 12.2 of Belton–Guillot–Khare–Putinar in the product topology on R^Z.","tokens_in":10488,"tokens_out":886,"duration_ms":28030,"significance":"Both results are precise, self-contained answers to named open questions in the recent total-positivity literature. The Bessel theorem extends the integer-order STP result of Buchstaber–Glutsyuk to all nonnegative real orders by a transparent Sturm–Darboux argument rather than the original Hilbert-space flow; the density theorem supplies a clean product-topology resolution of the Belton et al. question while carefully disclaiming stronger topologies. The proofs rely only on classical tools (Darboux–Crum Wronskian identity, Wronskian/ECT criterion, Cauchy–Binet, ASWE representation, generalized Vandermonde) and verify all inductive hypotheses explicitly for the Bessel family. The isolation of reusable mechanisms (spectral Darboux, discrete Toeplitz smoothing) and the open Question 5.1 on asymptotic propagation add structural value beyond the two theorems.","major_comments":[],"minor_comments":[{"comment":"In the inductive display after “Fix 0≤s1<⋯<sm” (proof of Theorem 1.1), the leading-coefficient formula is correct, but a one-line remark that the empty product is 1 and that every factor (sj-sℓ) is positive for j>r+1 would make the sign bookkeeping fully self-contained for a reader who skips the surrounding prose.","section":null},{"comment":"Lemma 4.2 invokes the ASWE Laurent representation for the non-geometric case; a parenthetical pointer to the precise form of the annulus (or to [2, §9]) would help readers who are not specialists in classical PF-sequence theory.","section":null},{"comment":"Section 5 notes that positive y-Wronskians are equivalent to positive ordinary x-Wronskians via the triangular change of derivative bases. Adding the explicit triangular factor x^{k(k-1)/2} already written in the text into a short displayed identity would make the Chebyshev-system corollary easier to cite.","section":null},{"comment":"The product-topology disclaimer is stated clearly in the abstract, introduction, and Remark 4.3; a single sentence in the statement of Theorem 1.2 itself (“in the product topology on R^Z”) would further reduce the chance of mis-citation as a norm-density result.","section":null}],"recommendation":"accept","confidential_remarks":"The manuscript is a clean fit for a classical-analysis / total-positivity venue: two short, independent answers to named open questions, both using only standard tools and with all inductive hypotheses verified in place. The unusually detailed generative-AI declaration is transparent and does not affect the mathematical content. I see no load-bearing gaps; an accept is appropriate."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This paper does exactly what it claims: it settles the real-order modified-Bessel STP question left open by Buchstaber–Glutsyuk and gives a product-topology answer to Belton–Guillot–Khare–Putinar’s density Question 12.2. Both proofs are short, self-contained, and use only standard tools (Darboux–Crum, Wronskian/ECT, Cauchy–Binet, ASWE, generalized Vandermonde).\n\nWhat is new is the real-order extension for K(x,s)=Is(x) on (0,∞)×[0,∞) and the explicit density statement that every two-sided PF sequence is a pointwise limit of STP ones. The integer-order Bessel result was already known; the contribution is the spectral-Darboux route that treats the order as a spectral parameter after the log change of variables. The density argument is likewise new in the form asked for: discrete-Gaussian Toeplitz smoothing plus a separate geometric-tilt case, with the product topology carefully scoped (no norm claim).\n\nThe soft spot the reader flagged—the inductive survival of left-endpoint asymptotics under Darboux stages—is real but not a gap. The paper verifies it explicitly for the Bessel family from the power series, termwise differentiation, and the inductive display of leading coefficients (positive products of (sj−sℓ)). Proposition 3.1 then applies directly. The Toeplitz half is classical and clean. Question 5.1 about general spectral families is left open and does not affect the theorems.\n\nCitations are appropriate; the arguments are independent of the target statements. No data, no free parameters, no circularity. This is for people who work on total positivity, Chebyshev systems, or PF sequences. It is useful, transparent, and short enough that a referee can check every line.\n\nI would send it to peer review without hesitation. Engage with it if either of those two questions sits in your orbit.","headline":"Clean answers to two published open questions via short classical arguments; the Bessel real-order extension and product-topology PF density both hold up.","tokens_in":11065,"tokens_out":503,"would_cite":true,"duration_ms":4357,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["15B48","33C10","34B24","30B10"],"pacs":[],"model":"grok-4.5","headline":"The modified-Bessel kernel is strictly totally positive for all nonnegative real orders, and totally positive Pólya-frequency sequences are dense in the product topology.","keywords":["total positivity","modified Bessel functions","Pólya-frequency sequences","Darboux transformation","Toeplitz kernels","strict total positivity","Wronskians"],"falsifier":"Compute the successive Darboux transforms of a concrete finite family of modified-Bessel functions (for example orders 0, 1/2, 1) and check whether the claimed leading asymptotic form and the positivity of all intermediate Wronskians hold numerically on a large negative half-line.","tokens_in":11234,"feed_emoji":"∫","tokens_out":702,"duration_ms":5670,"temperature":0.7,"pith_summary":"The paper isolates two mechanisms that force ordinary total nonnegativity into strict total positivity of infinite order. First, a spectral Darboux induction turns positivity plus ordered left-endpoint asymptotics of a one-parameter Sturm family into positive Wronskians and therefore into strictly positive evaluation determinants. Applied after the change of variables x = e^y, this shows that the modified-Bessel kernel K(x,s) = I_s(x) is strictly totally positive of infinite order for positive arguments and all nonnegative real orders, answering an open question left by the earlier integer-order theorem. Second, convolution against a discrete Gaussian Toeplitz kernel (or a multiplicative Gaussian tilt for pure geometrics) converts any two-sided Pólya-frequency sequence into a sequence whose Toeplitz kernel is strictly totally positive, while recovering the original sequence pointwise as the Gaussian width tends to zero. The two results therefore supply both a concrete strictification theorem for a classical special-function kernel and a product-topology density statement for discrete totally positive sequences.","feed_headline":"Bessel kernel is strictly totally positive for real orders","feed_subtitle":"A spectral Darboux induction settles the real-order determinant question and densifies PF sequences","key_machinery":"Spectral Darboux strictification (Proposition 3.1): an inductive chain of Darboux transforms that preserves positivity of the spectral solutions and propagates ordered left-endpoint asymptotics, thereby producing positive Wronskians that convert into positive evaluation determinants via the extended-Chebyshev criterion.","core_discovery":"The modified-Bessel kernel with real orders is strictly totally positive of infinite order on (0,∞)×[0,∞), and every two-sided Pólya-frequency sequence is a pointwise limit of totally positive Pólya-frequency sequences. Both statements are obtained by exhibiting explicit strictification mechanisms—spectral Darboux transport for the continuous kernel and discrete Gaussian Toeplitz smoothing for the sequences.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Spectral Darboux yields strict total positivity of modified-Bessel kernel for real orders","Bessel kernel I_s(x) strictly totally positive of infinite order via spectral induction","Toeplitz smoothing densifies every two-sided Pólya-frequency sequence pointwise","Strict TP from spectral Darboux transport and discrete Gaussian Toeplitz smoothing","Real-order Bessel determinants positive by Darboux; PF sequences densified in product topo"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The ordered left-endpoint asymptotics of the spectral solutions must survive every successive Darboux stage; if they fail to propagate, the induction no longer yields positive Wronskians.","fun_headline_variants_meta":{"raw":{"variants":["Spectral Darboux yields strict total positivity of modified-Bessel kernel for real orders","Bessel kernel I_s(x) strictly totally positive of infinite order via spectral induction","Toeplitz smoothing densifies every two-sided Pólya-frequency sequence pointwise","Strict TP from spectral Darboux transport and discrete Gaussian Toeplitz smoothing","Real-order Bessel determinants positive by Darboux; PF sequences densified in product topology"]},"model":"grok-4.5","effort":"low","cost_usd":0.005452,"raw_usage":{"total_tokens":1472,"prompt_tokens":749,"num_sources_used":0,"completion_tokens":114,"cost_in_usd_ticks":54520000,"prompt_tokens_details":{"text_tokens":749,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":609,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":749,"tokens_out":114,"duration_ms":5219,"temperature":1.0,"reasoning_tokens":609,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-12T07:05:09.560240+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Compute the successive Darboux transforms of a concrete finite family of modified-Bessel functions (for example orders 0, 1/2, 1) and check whether the claimed leading asymptotic form and the positivity of all intermediate Wronskians hold numerically on a large negative half-line.","supporting_citations":[],"review_version":1}