{"id":"ce11307d-a41e-43a7-b8e8-8d28a3f48714","arxiv_id":"2608.08198","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"The Srivastava-Tomovski function and its Laplace-Wright realization are completely monotone exactly when α≤κ and κβ≥αγ, with explicit Wright and beta Bernstein measures.","lead":"The paper gives the exact condition, α≤κ and κβ≥αγ, under which the Srivastava-Tomovski generalized Mittag-Leffler function is completely monotone on the positive half-line, and identifies the representing measures as known Wright and beta laws. A generalist might read it because complete monotonicity means a response can be viewed as a mixture of exponential decays, which matters in fractional calculus, relaxation modeling, and probability.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"All-real Mellin identity (13) is the load-bearing step for the μ<0 exclusion; its proof relies on an unverified scope of Wright's negative-ray theorem, so an independent check is warranted.","rationale":"The reader's weakest assumption correctly identifies the all-real Mellin identity as the load-bearing premise for the μ<0 exclusion. The rest of the classification is supported by independent, largely self-contained arguments: sufficiency comes from explicit positive Wright and beta densities, the boundary α=κ is handled by exact beta/atomic representations, and the exclusions α>κ and (α=κ, β<γ) follow from moment-support asymptotics that do not depend on the negative-ν Mellin identity. Proposition 5.1 is the only step that needs (13) at negative ν, and its internal proof is careful: the integration-by-parts boundary terms are checked, the holomorphic continuation is justified, and the exceptional nonpositive-integer ν cases are treated separately. The residual risk is confined to the cited Wright negative-ray theorem's scope. Because the cited theorem is standard in the second-kind Wright literature and the paper's own derivation is plausible, I do not downgrade the reader's accept verdict. The proposed numerical check would convert the moderate confidence into high confidence, but the concern as stated does not require changing the verdict.","tokens_in":13515,"tokens_out":50024,"duration_ms":427398,"concrete_test":"Verify (13) independently at the two ordinates used in Proposition 5.1, z0=-μ/a and z=γ, for parameter triples (a,μ) = (1/2, -1/4), (2/3, -1), (1/2, -2), and (3/4, -3), using a Mellin-Barnes (Fox-H) evaluation of W_{-a,μ}(-t) and high-precision quadrature of the integral ∫ t^{z-1} W_{-a,μ}(-t) dt against Γ(z)/Γ(μ+az). If any relative error exceeds 1e-8, the Mellin-zero step (26) is not established; if all cases pass, the necessity proof for 0<α<κ, μ<0 is confirmed.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central theorem's necessity direction for the excluded interior region 0<α<κ, μ<0 rests entirely on Lemma 4.1's Mellin identity (13), evaluated at z0=-μ/a and at z=γ. Proposition 5.1 uses the zero ordinate Γ(z0)/Γ(0)=0 to force the Wright kernel to change sign, and the positive ordinate at z=γ to show it also takes positive values. If (13) fails for some negative non-integer ν, the Mellin-zero step (26) does not hold, the sign-changing conclusion collapses, and the finite signed Laplace uniqueness argument no longer excludes a positive Bernstein measure. Appendix A proves (13) by invoking Wright's negative-ray theorem for arbitrary real ν [30] and then extending by N-fold integration by parts from q≥0 to negative ν. The algebraic and exponential bound (14) is needed uniformly for the auxiliary parameters ν+(j+1)a, j=0,...,N-1, some of which are negative. The paper cites [30,17] for this theorem but does not reproduce its hypotheses; if the bound or the removal of exceptional gamma-pole cases has an unstated restriction, the extension to all real ν is not established. This is a genuine correctness risk, not a disagreement with consensus.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies complete monotonicity of the Srivastava-Tomovski function E^{γ,κ}_{α,β}(-x) on the positive half-line for positive parameters. It carefully distinguishes the historical entire series, which is entire only when Δ = 1 + α - κ > 0, from a separately defined positive-axis Laplace-Wright realization that is defined for arbitrary positive parameters. The main results, Theorems 3.1 and 3.2, assert that complete monotonicity holds if and only if α ≤ κ and κβ ≥ αγ, both for the entire series in its entire regime and for the realization on the full positive parameter space. Sufficiency is established by explicit Bernstein measures: a strictly positive Wright density in the interior 0 < α < κ with μ = β - αγ/κ ≥ 0, a powered beta density at α = κ and β > γ, and an atom at α = κ and β = γ. Necessity uses three separate mechanisms: a Mellin-zero sign-changing argument for 0 < α < κ with μ < 0 (Proposition 5.1), a moment-support argument for α = κ with β < γ (Proposition 5.2), and a moment-support collapse for α > κ (Proposition 5.3). The appendices provide the all-real Mellin identity for the second-kind Wright function, the convergence trichotomy, finite signed Laplace uniqueness, and the moment-support lemma. The Prabhakar specialization κ = 1 is derived as Corollary 6.1.","tokens_in":13862,"tokens_out":21874,"duration_ms":179345,"significance":"If correct, the paper settles the complete-monotonicity classification for the four-parameter Srivastava-Tomovski function, going beyond known sufficiency results for the three-parameter Prabhakar function. The characterization is sharp, and the Bernstein measures are explicitly identified, including the atomic endpoint. The paper is unusually transparent about which ingredients are new and which are prior results: the interior positive measure is exactly Wang's power-biased Wright law, and the beta and atomic endpoints are inherited from Ferreira-Simon and related constructions. The proof structure is rigorous: the load-bearing Mellin identity for an arbitrary real secondary parameter is proved in Appendix A, and the excluded-region arguments rely on standard uniqueness and moment-support lemmas proved in Appendix C. The paper also delivers a clean consolidation of the Prabhakar case. I find the derivations sound and the claims falsifiable; the main residual risk, the scope of the external Wright negative-ray theorem, is addressed by explicit citation and by a self-contained integration-by-parts extension. There is no circular reasoning and there are no fitted parameters.","major_comments":[],"minor_comments":[{"comment":"The notation for the series and for the realization should be made visually distinct; in the current typesetting the symbols E and E are easy to confuse, which matters because the two objects are not equal in the zero-radius phase.","section":"Definition 2.3 and throughout"},{"comment":"The text refers to 'Theorem 4.1' in the proof of Proposition 4.2, whereas the result is labeled Lemma 4.1 in the main text; the numbering should be harmonized between the main text and the appendices.","section":"Section 4, proof of Proposition 4.2"},{"comment":"Please state explicitly that the constants C and N in the bound (14) may depend on ν but that, for the finite family of shifted parameters ν+(j+1)a appearing in the integration-by-parts boundary terms, they can be chosen uniformly; this would remove a small potential ambiguity in the boundary-term argument.","section":"Appendix A, after Eq. (34)"},{"comment":"The proof that W_{-a,μ}(-t) is not identically zero is correct, but the reasoning would be easier to follow if the coefficient conditions for n=0 and n=1 were written out explicitly, since this step is essential for the sign-changing conclusion.","section":"Proposition 5.1"}],"recommendation":"accept","confidential_remarks":"The manuscript relies on several arXiv preprints ([22], [25], [28], [29]) for load-bearing prior results. I have checked that the essential positivity input is also backed by the peer-reviewed Ferreira-Simon paper, and the preprint results are cited with theorem numbers. This does not affect my recommendation. The paper fits the journal's scope, and the probabilistic interpretation of the explicit measures is a genuine strength."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nTwo things to know up front. First, this paper gives the first if-and-only-if complete-monotonicity classification for the four-parameter Srivastava-Tomovski function: in the entire-series regime (Delta>0) and for the separately defined Laplace-Wright realization, complete monotonicity holds exactly when alpha<=kappa and kappa*beta>=alpha*gamma, and the Bernstein measures are explicit in every admissible case. Second, the advance is mostly synthesis, but it is careful synthesis with one genuinely load-bearing new piece: the Mellin identity for the second-kind Wright function with arbitrary real nu. That identity is also the place to look for trouble.\n\nWhat the paper does well: it keeps the generalized-Wright series separate from the positive-axis Laplace-Wright realization across the entire/finite-radius/zero-radius trichotomy, which matters because the series is not analytic in the zero-radius phase. The sufficiency direction is constructive: the interior measure is exactly Wang's power-biased Wright law pushed forward by t -> t^kappa, with Ferreira-Simon supplying positivity, moments, and endpoints. Necessity is split into three mechanisms - Mellin-zero for mu<0, moment-root support for beta<gamma, support-collapse for alpha>kappa - each clean, each backed by worked examples. The authors are explicit about what is prior and what is new; no fitted constants, no self-citation games, low circularity burden.\n\nSoft spots, in proportion. The main residual risk is the one any careful reader will flag: the mu<0 exclusion rests on the Mellin identity (13) for arbitrary real nu, proved in Appendix A by invoking Wright's negative-ray theorem and extending by integration by parts. The paper cites [30,17] but does not reproduce the theorem's hypotheses. The uniform bound (14) is asserted for auxiliary parameters nu+(j+1)a, some negative, and an unstated restriction there could break the sign-changing kernel conclusion. The authors do handle the exceptional nu in nonpositive integers explicitly, which shows awareness, but they do not quote the exact scope of Wright's theorem. I do not see an actual gap in their derivation - the integration by parts is spelled out - but this is a genuine correctness question for a referee to chase. Minor: several load-bearing inputs are arXiv preprints (Wang, Salazar, Sibisi), which is not a flaw but does temper confidence.\n\nBottom line: a solid, honest paper with a real, though moderate, advance. It deserves a serious referee, and I would take the referee report seriously. Send it out and ask the referee to verify Wright's negative-ray hypotheses and the uniform bound in Appendix A.","headline":"A careful synthesis paper that delivers the first if-and-only-if complete-monotonicity classification for the four-parameter Srivastava-Tomovski function, with the main residual risk isolated in the Mellin identity for the second-kind Wright function with negative parameter.","tokens_in":14262,"tokens_out":2413,"would_cite":true,"duration_ms":21793,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["33E12","26A48","44A10","33C60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves an exact if-and-only-if criterion for the Srivastava–Tomovski function to be completely monotone, and identifies the representing probability measure explicitly.","keywords":["complete monotonicity","Srivastava–Tomovski function","Prabhakar function","Wright function","Bernstein measure","generalized Wright series","Laplace transform","Mittag–Leffler function"],"falsifier":"Take $\\alpha=1$, $\\kappa=2$, $\\gamma=2$, $\\beta=1/2$, so $a=1/2$ and $\\mu=-1/2$. The central Mellin identity (13) predicts $\\int_0^\\infty W_{-1/2,-1/2}(-t)\\,dt=0$, since the right-hand side contains $\\Gamma(0)$ in the denominator. Evaluating this integral by high-precision numerical quadrature, using the Wright series near zero and its stretched-exponential asymptotics at infinity, either confirms zero, as the theorem requires, or returns a nonzero value, which would refute the identity and the necessity argument.","tokens_in":13319,"feed_emoji":"📉","tokens_out":11254,"duration_ms":88421,"temperature":0.7,"pith_summary":"This paper fixes exactly when the Srivastava–Tomovski function, a four-parameter Mittag–Leffler generalization, is completely monotone on the positive half-line: every derivative alternates in sign, or equivalently the function is the Laplace transform of a positive measure. For positive parameters $\\alpha,\\beta,\\gamma,\\kappa$, the negative-axis series is proved completely monotone if and only if $\\alpha\\le\\kappa$ and $\\kappa\\beta\\ge\\alpha\\gamma$, in the entire-series regime $\\Delta=1+\\alpha-\\kappa>0$. A separately defined Laplace–Wright realization, which exists for all positive parameters, obeys the same characterization. In every admissible case the representing Bernstein measure is explicit: a strictly positive Wright density, a powered $\\beta$ density, or a point mass, each with total mass $1/\\Gamma(\\beta)$. Because completely monotone functions are mixtures of exponential relaxations, the result identifies the exact parameter region in which this generalized relaxation kernel has a genuine positive spectral measure.","feed_headline":"Srivastava–Tomovski function: complete monotonicity fully classified","feed_subtitle":"The negative-axis series is a positive exponential mixture exactly when α≤κ and κβ≥αγ; the mixing law is explicit.","key_machinery":"The load-bearing object is the second-kind Wright function $W_{-a,\\nu}$ and its Mellin identity (13): $\\int_0^\\infty t^{z-1}W_{-a,\\nu}(-t)\\,dt=\\Gamma(z)/\\Gamma(\\nu+az)$ on the appropriate half-plane. The paper proves this identity for arbitrary real $\\nu$ by combining the stretched-exponential decay of $W_{-a,\\nu}$ with repeated integration by parts; this extension is what allows the kernel $K(u)=u^{\\gamma/\\kappa-1}W_{-\\alpha/\\kappa,\\mu}(-u^{1/\\kappa})/(\\kappa\\Gamma(\\gamma))$ to be analyzed even when $\\mu<0$. The sufficiency direction rides on the pushforward identity $\\Gamma(\\beta)K(u)\\,du=(t\\mapsto t^\\kappa)_\\# f_{a,\\mu,\\gamma-1}(t)\\,dt$, showing the normalized interior measure is a power-biased Wright law. The necessity direction uses two auxiliary tools: a finite signed Laplace uniqueness lemma (a signed measure with zero Laplace transform is zero) and a moment-root lemma that identifies the upper endpoint of a measure's support as the limit of $m_n^{1/n}$.","core_discovery":"The paper's central discovery is a complete positivity classification. Theorem 3.1 states that for $\\alpha,\\beta,\\gamma,\\kappa>0$ with $\\Delta=1+\\alpha-\\kappa>0$, the negative-axis Srivastava–Tomovski series $E^{\\gamma,\\kappa}_{\\alpha,\\beta}(-x)$ is completely monotone on $(0,\\infty)$ if and only if $\\alpha\\le\\kappa$ and $\\kappa\\beta\\ge\\alpha\\gamma$. Theorem 3.2 extends the same equivalence to the separately defined Laplace–Wright realization for arbitrary positive parameters, with strict complete monotonicity in every admissible case. The sufficiency direction identifies, up to normalization and the substitution $u=t^\\kappa$, the integrating kernel with a power-biased Wright law, giving an explicit positive measure; the necessity direction splits into three exclusions: for $0<\\alpha<\\kappa$ and $\\mu=\\beta-\\alpha\\gamma/\\kappa<0$, a zero in the Mellin transform forces the inverse kernel to change sign, so no positive measure can share its Laplace transform; for $\\alpha=\\kappa$ and $\\beta<\\gamma$, the moment roots force support in $[0,1]$ while the moments diverge; and for $\\alpha>\\kappa$, the moment roots collapse to zero, forcing support at a point while the first moment is positive. Every admissible Bernstein measure has total mass $1/\\Gamma(\\beta)$, so its normalization is a probability law.","pith_inferences":["The same mechanism — a zero ordinate in a Mellin transform forcing a sign-changing inverse — may classify complete monotonicity for other generalized Wright or Fox–H kernels, not just this four-parameter family.","Because the realization exists and is smooth but non-analytic at zero when $\\Delta<0$, this family offers explicit completely monotone functions that are not determined by a convergent Taylor series, which could be a testbed for numerical Laplace inversion.","The explicit probability structure suggests Bayesian or simulation interpretations: the normalized kernel is a prior on $(0,\\infty)$ with tunable tail behavior, with the atomic endpoint as a degenerate limiting law.","One could test whether the same inequalities characterize fractional complete monotonicity or Bernstein-function properties of the realization, questions the paper does not address."],"forward_implications":["For the Prabhakar specialization $\\kappa=1$, the theorem becomes: $E^\\gamma_{\\alpha,\\beta}(-x)$ is completely monotone if and only if $0<\\alpha\\le 1$ and $\\beta\\ge\\alpha\\gamma$, unifying earlier sufficiency and negative-value results.","In every admissible case the normalized Bernstein measure is a probability law, so the relaxation kernel can be evaluated or sampled by drawing from an explicit Wright, beta, or atomic law and exponentiating the draw.","Fractional-calculus operators whose kernels are built from this function inherit positivity, as Laplace transforms of positive measures, exactly in the parameter region $\\alpha\\le\\kappa$ and $\\kappa\\beta\\ge\\alpha\\gamma$.","The theorem separates the analytic transition controlled by $\\Delta=1+\\alpha-\\kappa$ from the positivity transition controlled by $\\alpha\\le\\kappa$ and $\\mu\\ge 0$; neither condition subsumes the other.","At the boundary $\\alpha=\\kappa$ the spectrum changes type, from a Wright density when $\\mu\\ge0$ to a compactly supported powered-beta law when $\\beta>\\gamma$, and finally to the point mass $e^{-x}/\\Gamma(\\gamma)$ when $\\beta=\\gamma$."],"supporting_citations":[{"why":"Supplies the second-kind Wright density's positivity, admissible parameter region, moments, and beta/atomic endpoint behavior, which form the interior measure.","marker":"[2]"},{"why":"Introduces the Srivastava–Tomovski series and the generalized-Wright convergence criterion that determines the entire/finite-radius/zero-radius trichotomy.","marker":"[26]"},{"why":"Proves the Prabhakar sufficiency statement that the present theorem extends to an if-and-only-if classification.","marker":"[8]"},{"why":"Provides the power-biased Wright law serving as the normalized interior measure and the negative-value result covering $\\beta<\\alpha\\gamma$ in the Prabhakar specialization.","marker":"[28]"},{"why":"Supplies the negative-ray asymptotics behind the stretched-exponential bound on which the Mellin identity and integration by parts rely.","marker":"[30]"},{"why":"Gives the beta-power law that appears at the equal-step boundary $\\alpha=\\kappa$, $\\beta>\\gamma$.","marker":"[9]"},{"why":"Provides the zero-balanced atomic endpoint and weak convergence mechanism used for $\\beta=\\gamma$.","marker":"[10]"},{"why":"States Bernstein's theorem and the Laplace uniqueness for positive measures that connect complete monotonicity to a positive spectrum.","marker":"[23]"}],"fun_headline_variants":["Srivastava–Tomovski positive mixture iff α≤κ and κβ≥αγ","When is Srivastava–Tomovski completely monotone?","Explicit Bernstein law for Srivastava–Tomovski","Srivastava–Tomovski: the exact monotonicity condition"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the all-real Mellin identity for the second-kind Wright function: the integral $\\int_0^\\infty t^{z-1}W_{-a,\\nu}(-t)\\,dt$ must equal $\\Gamma(z)/\\Gamma(\\nu+az)$ for every real $\\nu$, including negative values, with no hidden exceptional case. If that identity failed for some negative $\\nu$, the proof that the region $0<\\alpha<\\kappa$, $\\beta<\\alpha\\gamma/\\kappa$ is not completely monotone would collapse, and the necessity direction of the main theorem with it.","fun_headline_variants_meta":{"raw":{"variants":["Srivastava–Tomovski positive mixture iff α≤κ and κβ≥αγ","When is Srivastava–Tomovski completely monotone?","Explicit Bernstein law for Srivastava–Tomovski","Srivastava–Tomovski: the exact monotonicity condition"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000624,"raw_usage":{"total_tokens":2959,"prompt_tokens":1085,"completion_tokens":1874,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":701,"completion_tokens_details":{"reasoning_tokens":1789}},"tokens_in":701,"tokens_out":1874,"duration_ms":14558,"temperature":1.0,"reasoning_tokens":1789,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T14:27:37.046670+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take $\\alpha=1$, $\\kappa=2$, $\\gamma=2$, $\\beta=1/2$, so $a=1/2$ and $\\mu=-1/2$. The central Mellin identity (13) predicts $\\int_0^\\infty W_{-1/2,-1/2}(-t)\\,dt=0$, since the right-hand side contains $\\Gamma(0)$ in the denominator. Evaluating this integral by high-precision numerical quadrature, using the Wright series near zero and its stretched-exponential asymptotics at infinity, either confirms zero, as the theorem requires, or returns a nonzero value, which would refute the identity and the necessity argument.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the beta-power law that appears at the equal-step boundary $\\alpha=\\kappa$, $\\beta>\\gamma$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the zero-balanced atomic endpoint and weak convergence mechanism used for $\\beta=\\gamma$."}],"review_version":2}