{"id":"b4e6f14f-ed47-4903-a673-4bbfada66c49","arxiv_id":"2507.22425","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Finite sums of consecutive Laguerre polynomials are real-rooted for large n exactly when an auxiliary coefficient polynomial has only real roots, with four Laguerre normalizations giving different counts of non-real zeros.","lead":"The paper proves that the number of real zeros of sums of K+1 consecutive Laguerre polynomials is controlled, for large n, by the roots of a single auxiliary polynomial that depends only on the coefficients. This gives a nearly complete, normalization-dependent real-rootedness theory for these finite combinations, with explicit thresholds and interlacing results.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Lemma 3.4's bridge identity q_n = (-1)^n b^{α+n-K+1;φ}_n is proved only for n=K and n=K+1; the text says the remaining cases follow 'similarly' without giving the induction. Since Theorem 1.3 uses this identity for every n ≥ K, the monic-case theorem rests on an omitted proof.","rationale":"The reader's weakest assumption is indeed the most load-bearing concern. Theorems 1.3 through 1.6 are the central results, and Theorem 1.3 depends on Lemma 3.4 to transfer zero information from generalized Bell polynomials to the monic Laguerre combinations. The submitted text explicitly leaves the induction for n > K+1 to 'a similar argument,' which is an omitted proof rather than a mere stylistic shortcut, because no recurrence relating the n-level and (n+1)-level identity is written down. This does not force rejection: the identity is consistent with the proved cases and with the surrounding quasi-spectral machinery, and the remaining theorems (1.4, 1.5, 1.6) do not rely on this lemma. The reader's CONDITIONAL verdict already captures the situation exactly: the central claims are credible and likely correct, but the manuscript is not fully self-contained as submitted. The secondary issue in Remark 5.2 is a real internal inconsistency, but it concerns a corollary rather than the main theorems. Therefore no change to the reader's verdict is needed.","tokens_in":33103,"tokens_out":4620,"duration_ms":54245,"concrete_test":"Have the author supply the missing induction in Lemma 3.4, or independently verify it: for K=3,4 and generic real zeros θ_i of Q (for instance θ_i = i, i=1,...,K), compute both sides of (3.11) for n=K+2, K+3, K+4 in exact arithmetic, using (3.3) for q_n and recurrence (3.5) for b^{α+n-K+1;φ}_n, and compare coefficients. Also resolve the operator-index inconsistency in the n=K+1 step by stating explicitly whether Λ_{α+1} or Λ_{α+2} is applied, and recheck the last sentence of Remark 5.2 against the inequality it derives.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing gap is in Lemma 3.4, after equation (3.17). The proof handles the base case NA=0, then proves the identity for n=K and n=K+1, and ends with 'The proof for the rest of the cases can be completed proceeding similarly.' But Theorem 1.3, both in Step 1 and in Step 2 via (3.23), invokes (3.11) for all n ≥ K. If the omitted induction fails for some n > K+1, the step from the generalized Bell polynomials back to the original Laguerre combinations is broken, and the entire monic theorem loses its proof. The identity is plausible and likely provable by induction using (3.5) and (3.2), but as written it is not demonstrated. The same passage also contains a small operator-index slip: the text says 'apply Λ_{α+2}' and then writes Λ_{α+1} in the following display, which should be corrected when the induction is supplied. In addition, Remark 5.2 contains an apparent self-contradiction: after proving that Corollary 5.4 cannot be true for noninteger α < K-2, it concludes 'and so Corollary 5.4 can be true.' That error affects Corollary 1.7(1) and should be fixed, but it is secondary to the Lemma 3.4 gap.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the real and non-real zeros of finite linear combinations of K+1 consecutive Laguerre polynomials in four normalizations: monic, value-one-at-zero, standard, and Brenke. The main results, Theorems 1.3–1.6, assert that for large n the zero behaviour of q_n is governed by the auxiliary polynomial Q(x)=Σ(-1)^j γ_j (x)_{K-j} in the monic case and by P(x)=Σ γ_j x^{K-j} in the other three cases. In particular, the paper gives conditions on the zeros of Q or P under which all zeros of q_n are real, simple, positive, or interlace, and, when P has non-real zeros, it describes exactly how many non-real zeros persist for large n. The proofs combine the quasi-spectral properties of Laguerre polynomials with generalized Bell-type polynomial families, complex-zero-decreasing operators, and asymptotic expansions.","tokens_in":33489,"tokens_out":5492,"duration_ms":59372,"significance":"If the main theorems are correct, this is a substantial contribution to the zero-location theory of Laguerre combinations and to the real-rootedness preservation literature. The paper improves earlier results of Iserles, Nørsett, and Saff, provides explicit thresholds n_0 and n_1, gives a counterexample showing that interlacing can fail for small n, and connects the problem to generalized Bell polynomials and Brenke polynomials. The four-normalization comparison is a useful unifying framework. The paper is generally detailed and mostly self-contained, and the statements of Theorems 1.3–1.6 are sharp and falsifiable.","major_comments":[{"comment":"The bridge identity (3.11) is proved only for the case NA=0 and then for n=K and n=K+1; the text states that 'the proof for the rest of the cases can be completed proceeding similarly.' Since Theorem 1.3 uses (3.11) for every n ≥ K in Step 1 and uses (3.23) in Step 2, the monic-case theorem rests on an omitted induction. Please supply the full induction step for n ≥ K+2, or provide a precise reduction to the proved cases. The operator-index slip in this passage ('apply Λ_{α+2}' followed by Λ_{α+1} in the display) should also be corrected.","section":"Section 3, Lemma 3.4"},{"comment":"Remark 5.2 is self-contradictory. It aims to show that Corollary 5.4 cannot be true for noninteger α < K-2, but after using [29, Theorem 6.73] to bound the number of real zeros of q_n by n + floor(α-K+1) + 1, it concludes 'and so Corollary 5.4 can be true.' The displayed inequality is also wrong: for noninteger α < K-2 one has floor(α-K+1)+1 ≤ -1, so q_n has fewer than n real zeros. The correct conclusion is that Corollary 5.4 cannot hold. This error affects Corollary 1.7(1) and the surrounding discussion.","section":"Section 5, Remark 5.2"},{"comment":"Lemma 2.6 is stated with the proof omitted ('the proof is similar to the usual proof for Hurwitz's Theorem'). The lemma is used as a load-bearing tool in Step 6 of the proof of Theorem 1.4 and again in the interlacing argument in Theorem 1.5. Since it guarantees uniform persistence of non-real zeros, the proof or a precise reference should be included rather than left to the reader.","section":"Section 2, Lemma 2.6"}],"minor_comments":[{"comment":"The proof of Lemma 5.2 is omitted with the explanation that it is the same as that of Lemma 4.3. This is acceptable only if the identical calculation is explicitly acknowledged; please include the statement or a precise pointer so the reader does not need to reconstruct the asymptotic.","section":"Section 5, Lemma 5.2"},{"comment":"The proof of Corollary 8.2 is omitted because it is similar to that of Corollary 8.1. Since this corollary is not central to the main theorems, a clear reference to Corollary 8.1 is enough, but the sentence should be completed with the relevant details or a citation.","section":"Section 8, Corollary 8.2"},{"comment":"The abstract states 'if P has m>1 non-real zeros', but the theorems treat the case N_nr > 0, i.e., m ≥ 1. Please correct the inequality.","section":"Abstract and Section 1"},{"comment":"The phrase 'second, third and forth cases' contains a typo; it should be 'fourth'.","section":"Abstract"},{"comment":"The notation in Theorem 1.3 uses m for the number of real zeros satisfying α+1 ≤ θ_j, while the proof of Step 1 uses r_1 for the same quantity. Please make the notation consistent to avoid confusion.","section":"Section 3, Theorem 1.3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript depends heavily on the author's previous preprints [8] and [9]. The editor may wish to verify that these are publicly available or in press, since Corollary 1.2 and the Brenke results in Theorem 1.6 are used as black boxes. The main technical issue is the omitted induction in Lemma 3.4; this is likely repairable, and the rest of the paper is detailed enough that I would not recommend rejection on that basis alone."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The main results are a real advance. For four normalizations of Laguerre polynomials, the paper gives a nearly complete description of the real and non-real zeros of q_n = sum γ_j L^α_{n-j}, with explicit thresholds n0, n1, interlacing, and asymptotic zero distributions. The key insight — that the auxiliary polynomials Q(x)=sum (-1)^j γ_j (x)_{K-j} and P(x)=sum γ_j x^{K-j} control the story — is well supported. The monic case sharpens the author's own Corollary 1.2, the value-1-at-0 and Brenke cases are genuinely new, and the Bell-polynomial representation is a clever proof device. The paper also does a service by comparing with Iserles–Nørsett–Saff and showing how normalization changes the answer.\n\nThe soft spots are real but not fatal. The biggest is Lemma 3.4: the bridge identity q_n = (-1)^n b^{α+n-K+1;φ}_n is proved for n=K and n=K+1, and the rest is dismissed with \"proceeding similarly.\" Theorem 1.3 uses this identity for every n ≥ K, so the monic theorem as written rests on an omitted induction. The identity is plausible and probably provable, but the text is not self-contained on this point. A referee should demand the missing argument or an explicit reduction to a lemma with a proof. There is also an operator-index typo in that same proof (apply Λ_{α+2}, then write Λ_{α+1}).\n\nSecond, Remark 5.2 contains a sign error: after showing that Corollary 5.4 cannot be true for noninteger α < K-2, it concludes \"and so Corollary 5.4 can be true.\" That contradicts the preceding argument and affects the statement of Corollary 1.7(1). It needs a straightforward fix. Minor omissions: Lemma 2.6, Lemma 5.2, and Corollary 8.2 are stated without proofs; the first is a routine Hurwitz-type stability fact, the other two are said to follow by the same calculation as Lemma 4.3 and Corollary 8.1. These are acceptable if the references are precise, but they should be checked.\n\nOverall, the mathematics looks coherent and the central claims are likely correct. This is a paper for specialists in orthogonal polynomials and zero location, and it deserves careful peer review — not a desk rejection. My recommendation: send it out, with the clear instruction that the Lemma 3.4 gap and Remark 5.2 sign error be resolved before publication.","headline":"A solid, substantial contribution to the zero theory of finite Laguerre sums; the main theorems look right, but the monic-case bridge identity is not fully proved as written and one remark contains a sign error.","tokens_in":33996,"tokens_out":2812,"would_cite":true,"duration_ms":30563,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["42C05","26C10","33C45"],"pacs":[],"model":"deepseek-v4-flash","headline":"A single auxiliary polynomial Q or P determines, for large n, how many real zeros a finite consecutive Laguerre combination has.","keywords":["zeros","Laguerre polynomials","linear combinations","real-rootedness","quasi-spectral properties","generalized Bell polynomials","interlacing","Brenke polynomials"],"falsifier":"Take the sharpest monic threshold: $\\alpha=0$, $K=2$, $\\theta_1=1/4$, $\\theta_2=1/2$, which via (3.18)–(3.19) gives $\\gamma_1=7/4$, $\\gamma_2=1/8$; Theorem 1.3(1) predicts that every $q_n$, starting at $n_0=2$, has only positive simple zeros, so computing the zeros of $q_2=\\hat L_2^0+\\frac74\\hat L_1^0+\\frac18\\hat L_0^0$, and then of $q_3$ and $q_4$, either confirms the asserted thresholds or exhibits a concrete failure of the bridge identity's omitted induction.","tokens_in":32861,"feed_emoji":"🧮","tokens_out":11818,"duration_ms":126557,"temperature":0.7,"pith_summary":"This paper establishes that for a finite combination $q_n(x)=\\sum_{j=0}^K \\gamma_j \\tilde L_{n-j}^\\alpha(x)$ of $K+1$ consecutive Laguerre polynomials, the eventual number and location of real zeros is read off from a single auxiliary polynomial: $Q(x)=\\sum_{j=0}^K(-1)^j\\gamma_j(x)_{K-j}$ for the monic normalization and $P(x)=\\sum_{j=0}^K\\gamma_j x^{K-j}$ for the other three normalizations. If $Q$ has only real zeros, all zeros of $q_n$ are real and simple for $n\\ge n_1$ and positive for $n\\ge n_0$, with explicit thresholds computed from the zeros of $Q$; if $Q$ has non-real zeros, the monic $q_n$ still has only positive simple zeros for large $n$. In the other three cases, if $P$ has only real zeros then $q_n$ has only real zeros, and if $P$ has $m$ non-real zeros then exactly $m$ non-real zeros persist for large $n$. The paper also shows which of the four normalizations preserve real-rootedness under the map $T(x^n)=p_n$, and it recovers and sharpens earlier $n=K$ criteria. These criteria matter because earlier results guaranteed only that such combinations eventually have many real zeros; here the whole zero configuration is settled by a finite polynomial and explicit degree bounds.","feed_headline":"One auxiliary polynomial rules Laguerre zeros","feed_subtitle":"For four normalizations, P or Q decides when every zero is real—and how many are not.","key_machinery":"The engine is a set of quasi-spectral identities: first- or second-order differential operators act on each normalized Laguerre family as one-step shifts with constant eigenvalue, for example $\\Lambda_\\alpha(\\hat L_n^\\alpha)=-\\hat L_{n+1}^{\\alpha-1}$ for the monic family, $\\Upsilon_\\alpha(\\mathcal L_n^\\alpha)=\\mathcal L_n^{\\alpha-1}$ for the value-one-at-zero family, and $\\Omega_\\alpha$ as the corresponding shift for the Brenke family. For the monic case, iterating the backward shift connects $q_n$ to generalized Bell polynomials $b_n^{r;\\phi}$ defined by the recurrence $b_{n+1}=\\Lambda_r b_n+\\phi_{n+1}b_n$, where $\\phi$ is built from the zeros of $Q$; the bridge identity $q_n(x)=(-1)^n b_n^{\\alpha+n-K+1;\\phi}(x)$ (Lemma 3.4) transfers zero-counting to a theorem about zeros of these Bell-type polynomials. For the other normalizations, the differential operators are complex zero decreasing, and Laguerre asymptotics expressed through the conformal map $\\varphi(z)=\\tfrac12(z-2+\\sqrt{z^2-4z})$ convert the limiting zero set of $q_n$ into the zero set of $P(-\\varphi(z))$.","core_discovery":"The central discovery is that the four normalizations separate into two regimes. For monic Laguerre polynomials, Theorem 1.3 proves that the zeros of $Q$ are the only data needed: if all zeros of $Q$ are real, every $q_n$ has real and simple zeros for $n\\ge n_1$ and positive and simple zeros for $n\\ge n_0$, with $n_l=\\max\\{K,\\lfloor\\theta_i-\\alpha+K\\rfloor:1\\le i\\le m-l\\}$, and if $Q$ has non-real zeros, positivity still holds for large $n$. For the normalized families, Theorems 1.4 through 1.6 prove that $P$ controls the count: under the disk condition $P(z)\\ne 0$ for non-real $|z|\\le 1$, if $P$ has $N^{\\rm nr}$ non-real zeros, then for large $n$ the polynomial $q_n$ has exactly $n-N^{\\rm nr}$ real simple zeros, with $N^1$ of them negative, and the real zeros of consecutive $q_n$ interlace. In the Brenke normalization the statement becomes an equivalence: $q_n$ has only real zeros for all $n$ if and only if all zeros of $P$ are real; otherwise exactly $N^{\\rm nr}$ non-real zeros persist. Thus normalization is not a cosmetic detail—it changes the answer, and the paper quantifies precisely how.","pith_inferences":["The same quasi-spectral mechanism should yield finite auxiliary polynomials for other classical or semi-classical families with an explicit backward-shift operator, giving exact thresholds rather than asymptotic statements.","The explicit $n_0$ and $n_1$ formulas turn the monic theorem into a finite algebraic certificate: test the auxiliary polynomial for real roots, then a finite computation validates all degrees.","The Brenke equivalence suggests that any failure of real-rootedness in that normalization must come from the non-real zeros of $P$, not from the Bessel-type factor in the generating function, which could guide searches for real-rooted generating functions.","The contrast between the monic and Brenke normalizations indicates that normalization can be chosen deliberately when a finite combination is engineered to have prescribed real-zero behaviour, for example in quadrature or spectral constructions."],"forward_implications":["For monic combinations with $Q$ having only real zeros, all zeros of $q_n$ are real and simple once $n$ reaches an explicit bound computed from the large zeros of $Q$; when all zeros of $Q$ lie below $\\alpha+1$, real-and-positive rootedness holds immediately for every $n\\ge K$.","For the value-one-at-zero and standard Laguerre normalizations, if $P$ has no non-real zeros in the closed unit disk and $P(1)\\ne 0$, the number of real zeros of $q_n$ is exactly $n-N^{\\rm nr}$ for large $n$, with exactly $N^1$ negative zeros, and the real zeros of consecutive $q_n$ interlace.","For the Brenke normalization, real-rootedness of $q_n$ for all $n$ is equivalent to $P$ having only real zeros; otherwise exactly $N^{\\rm nr}$ non-real zeros survive.","The $n=K$ case recovers and sharpens earlier criteria: when the auxiliary polynomial has only real zeros and the Laguerre parameter is in the stated range, expansions of the form $\\sum_j \\tau_j L_j^\\alpha$ are real-rooted.","Corollary 1.7 gives new cases in which the linear operator $T(x^n)=p_n$ preserves real-rootedness, including all real $\\alpha>-1$ for two of the normalizations and integer-$\\alpha$ cases for the standard normalization."],"supporting_citations":[{"why":"Supplies the predecessor result on eventual real-rootedness of monic combinations and Corollary 1.2, used in Steps 2 and 4 of Theorem 1.3.","marker":"[9]"},{"why":"Introduces generalized Bell polynomials whose zero properties are transferred to the monic case by the bridge identity Lemma 3.4.","marker":"[7]"},{"why":"Provides the theorem that at least $n-K$ zeros of $q_n$ lie in intervals between zeros of $L_n$, used in Theorems 1.5 and 1.6 and in the asymptotic corollaries.","marker":"[2]"},{"why":"Supplies the Laguerre asymptotics via the conformal map $\\varphi(z)$ that drive the zero-counting limits in the non-monic cases.","marker":"[14]"},{"why":"Gives the Brenke-polynomial machinery, including the generating function and the Laguerre-Pólya characterization of real-rootedness used in Theorem 1.6.","marker":"[8]"},{"why":"Provides the complex zero decreasing operator theory used to control real and non-real zero counts under the differential operators of Sections 4 and 6.","marker":"[5]"},{"why":"Establishes the earlier role of polynomials of the form $\\sum \\gamma_j x^j$ in real-rootedness of Laguerre expansions; serves as the comparison baseline for Section 7.","marker":"[18]"},{"why":"Supplies standard Laguerre zero-location facts and identities, including Theorem 6.31.2 and the result used in Remark 5.2.","marker":"[29]"}],"fun_headline_variants":["One polynomial decides all zeros of Laguerre sums","Normalization flips when Laguerre sums have real zeros","Auxiliary polynomial governs real zeros for four normalizations","Exact count of non-real zeros in Laguerre sums","Laguerre zero behavior encoded in one polynomial"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The monic theorem depends on the bridge identity $q_n(x)=(-1)^n b_n^{\\alpha+n-K+1;\\phi}(x)$ holding for every $n\\ge K$, but Lemma 3.4 verifies only $n=K$ and $n=K+1$ and asserts the rest \"can be completed proceeding similarly\"; if that omitted induction has a hidden case-dependence, Theorem 1.3 collapses, and the non-monic theorems additionally require the disk condition $P(z)\\ne 0$ for non-real $|z|\\le 1$.","fun_headline_variants_meta":{"raw":{"variants":["One polynomial decides all zeros of Laguerre sums","Normalization flips when Laguerre sums have real zeros","Auxiliary polynomial governs real zeros for four normalizations","Exact count of non-real zeros in Laguerre sums","Laguerre zero behavior encoded in one polynomial"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000753,"raw_usage":{"total_tokens":3489,"prompt_tokens":1225,"completion_tokens":2264,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":841,"completion_tokens_details":{"reasoning_tokens":2184}},"tokens_in":841,"tokens_out":2264,"duration_ms":18201,"temperature":1.0,"reasoning_tokens":2184,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T11:42:35.241886+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the sharpest monic threshold: $\\alpha=0$, $K=2$, $\\theta_1=1/4$, $\\theta_2=1/2$, which via (3.18)–(3.19) gives $\\gamma_1=7/4$, $\\gamma_2=1/8$; Theorem 1.3(1) predicts that every $q_n$, starting at $n_0=2$, has only positive simple zeros, so computing the zeros of $q_2=\\hat L_2^0+\\frac74\\hat L_1^0+\\frac18\\hat L_0^0$, and then of $q_3$ and $q_4$, either confirms the asserted thresholds or exhibits a concrete failure of the bridge identity's omitted induction.","supporting_citations":[{"cited_title":"Dur´ an, Generalized Bell polynomials, J","cited_arxiv_id":null,"evidence_quote":"Introduces generalized Bell polynomials whose zero properties are transferred to the monic case by the bridge identity Lemma 3.4."},{"cited_title":"Beardon, K.A","cited_arxiv_id":null,"evidence_quote":"Provides the theorem that at least $n-K$ zeros of $q_n$ lie in intervals between zeros of $L_n$, used in Theorems 1.5 and 1.6 and in the asymptotic corollaries."},{"cited_title":"Geronimo, W","cited_arxiv_id":null,"evidence_quote":"Supplies the Laguerre asymptotics via the conformal map $\\varphi(z)$ that drive the zero-counting limits in the non-monic cases."},{"cited_title":"Brenke polynomials with real zeros and the Riemann Hypothesis","cited_arxiv_id":"2405.18940","evidence_quote":"Gives the Brenke-polynomial machinery, including the generating function and the Laguerre-Pólya characterization of real-rootedness used in Theorem 1.6."},{"cited_title":"Craven and G","cited_arxiv_id":null,"evidence_quote":"Provides the complex zero decreasing operator theory used to control real and non-real zero counts under the differential operators of Sections 4 and 6."},{"cited_title":"Iserles and E","cited_arxiv_id":null,"evidence_quote":"Establishes the earlier role of polynomials of the form $\\sum \\gamma_j x^j$ in real-rootedness of Laguerre expansions; serves as the comparison baseline for Section 7."},{"cited_title":"Szeg¨ o, Orthogonal Polynomials","cited_arxiv_id":null,"evidence_quote":"Supplies standard Laguerre zero-location facts and identities, including Theorem 6.31.2 and the result used in Remark 5.2."}],"review_version":1}