{"id":"a657972d-70cb-4f3d-9b65-517fc04e90c1","arxiv_id":"2507.06716","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For σ in (0,1], the paper constructs an explicit optimal Hardy weight W^op_σ for (-Δ_N)^σ, with W^op_σ(n) approximately n^{-2σ} and an upper bound C_σ for the best constant in the n^{-2σ} Hardy inequality.","lead":"This paper finds the optimal Hardy weight for fractional powers of the discrete Laplacian on the positive integers, for every fractional exponent up to 1. It settles an open question about the sharp constant in the classical n^{-2σ} Hardy inequality and adds a unique-continuation result at infinity.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central proof applies [37]'s criticality theory to an infinite-range graph without verifying that its theorems do not require local finiteness; this conditions the optimality theorem.","rationale":"I agree with the reader that the most load-bearing assumption is the transfer of criticality theory to a non-locally-finite graph, not the internal kernel estimates. I checked the main internal arguments: Proposition 3.2's dominated-convergence step with the growth estimates (3.6), Proposition 3.6's energy computation for the null-sequence φ_k, Proposition 3.8's ℓ^2(N, W_α,σ) threshold, and Proposition 3.10's monotonicity of Ψ_σ and contradiction via Lemma 3.9 are all coherent and essentially correct for σ ∈ (0,1), with the σ=1 case handled by the Liouville comparison principle. The only genuine soft spot is the unverified applicability of [37] and [11] to a graph with infinite degree and summable edge weights. The text itself flags local finiteness as a technical requirement in the earlier optimal-weight construction, and no passage verifies that the null-sequence criterion, ground-state representation, and comparison principle remain valid without it. This is not an internal inconsistency, but it is a missing verification of a standing hypothesis. Since the reader already made the verdict conditional on this check, I do not change the verdict; a check of the cited hypotheses would settle whether the paper is essentially complete.","tokens_in":1,"tokens_out":23724,"duration_ms":244326,"concrete_test":"Inspect the hypotheses of [37, Theorems 4.2 and 5.3] and [11, Theorem 2.1] as published. If local finiteness is assumed, test the null-sequence criterion on the present graph by truncating the edge weights to −eK^σ_{m,n} 1_{|m−n|≤R}, proving the relevant criticality statements for each truncated locally finite graph, and showing the null-sequence energy estimates are uniform as R→∞; if the criterion extends to summable infinite-range weights without local finiteness, record that verification and the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The weakest point is an unstated structural hypothesis on the graph {N, −eK^σ, R^σ}. The Introduction explicitly says that local finiteness was required in the earlier optimal Hardy-weight construction [36]: \"in order to avoid certain technical difficulties in finding optimal Hardy-weights for graph Laplacian (Definition 2.1), we require the underlying graph to be locally finite.\" The fractional graph here is not locally finite: by Remark 2.3(iv), −eK^σ_{m,n} ≍ |m−n|^{−1−2σ} and is nonzero for every m ≠ n. Nevertheless, Propositions 2.5, 2.7, 2.8, and Theorem 2.11 are cited from [37] and [11] and used for this graph. Proposition 3.6 constructs a null-sequence via the ground-state representation (3.14) and the null-sequence criterion [37, Thm 5.3]; Proposition 3.10 uses Lemma 3.9 with the uniqueness/ground-state properties of Remark 2.8; Proposition 3.7 uses the Liouville-type comparison principle. If any of these theorems is stated for locally finite graphs only, the chain from Proposition 3.6 to Theorem 1.1 has a gap. The paper does not state or prove an extension to infinite-range, summable edge weights, even though its Definition 2.1 suggests such generality is intended. The optimality claim is therefore conditional on a plausible but unverified hypothesis.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the fractional Laplacian (-Δ_N)^σ on the discrete half-line N. Using the representation of this operator as a graph Laplacian for σ∈(0,1], the authors construct a family of positive Hardy weights W_{α,σ} = I_{α-σ}/I_α, where I_α is a Riesz potential defined in (3.1). They prove that W_{α,σ} is critical for (-Δ_N)^σ if and only if α ≤ (3+2σ)/4, and null-critical at α=(3+2σ)/4, yielding an explicit optimal Hardy weight W^op_σ with asymptotic n^{-2σ}; for σ=1 it is asymptotically larger than the Keller-Pinchover-Pogorzelski weight. A Landis-type unique continuation theorem for positive supersolutions is derived as an application.","tokens_in":24533,"tokens_out":25047,"duration_ms":267772,"significance":"The result is significant: it answers the optimal-weight question posed in [27] for the subcritical range σ∈(0,1], extends the integer-lattice construction of [34] to the half-line, determines the best Hardy constant at infinity for the weight n^{-2σ}, and gives a sharp unique continuation criterion. The proof is constructive and parameter-free: the weight is explicit in Gamma functions, the ground-state representation is derived directly from the kernel, and the criticality threshold is obtained from sharp summability estimates rather than from fitted parameters. These are concrete strengths. The main reservation is a missing verification of the structural hypotheses in the imported criticality theory.","major_comments":[{"comment":"The criticality arguments rely on Propositions 2.5, 2.7, Remark 2.8, and Theorem 2.11 imported from [37], [36], and [11]. For σ∈(0,1), the graph {N, -eK^σ, R^σ} is not locally finite: by Remark 2.3(iv), -eK^σ_{m,n} ≍ |m-n|^{-1-2σ} is nonzero for every m≠n. The Introduction (Section 1, page 3) states that local finiteness was imposed in [36] 'in order to avoid certain technical difficulties in finding optimal Hardy-weights for graph Laplacian,' and Remark 2.8 cites [36, Theorem 5.3] for the uniqueness of the Agmon ground state. The manuscript does not verify that the theorems of [37] and [11] used here remain valid for infinite-range, summable edge weights, nor that the fractional graph satisfies the standing assumptions of those papers. Since the null-sequence criterion (Proposition 2.7) drives Proposition 3.6 and the Liouville comparison principle (Theorem 2.11) drives Propositions 3.7 and Theorem 1.2, the optimality claim in Theorem 1.1 is conditional on an unstated structural hypothesis. The authors should either state explicitly that the cited criticality theory does not require local finiteness, with precise references, or prove the needed extension to infinite-range summable edge weights.","section":"Section 2.2 and the proofs of Propositions 3.6, 3.7, 3.10 and Theorem 1.2"}],"minor_comments":[{"comment":"The definition of the domain F_X contains a misprint: the summability condition should be Σ_{m∈X} b_{n,m}|f(m)| < ∞ for every n, not Σ_{m∈X} b_{n,m}|f(n)|.","section":"Definition 2.1"},{"comment":"The displayed formula for R^σ_n has a 0·∞ ambiguity at σ=1: the expression involving sin(πσ) is not literally well-defined there. The value R^1_n = δ_{1,n} should be obtained by a limiting argument or by a direct computation for σ=1.","section":"Section 2.1, Eq. (2.11)"},{"comment":"The simplified energy functional Q^σ_α is defined with sums over n,m∈Z, but I_α and K^σ are defined on N, and the proof of Proposition 3.6 sums over 1≤n<m. Please correct the index set or explicitly define the extensions to Z.","section":"Section 3.2, Eq. (3.13)"},{"comment":"The notation 'Qσ−α' in the proof of Lemma 3.9 is confusing and should be typeset as Q^σ_α (the subscript is α, not −α).","section":"Section 3.4, Lemma 3.9"},{"comment":"In the case analysis for (3.15), situation (i) writes '(2 − 1)²' where '(1 − 0)²' is meant, and situation (ii) contains the typo '=≤'. These should be corrected.","section":"Proposition 3.6, proof of (3.15)"}],"recommendation":"major_revision","confidential_remarks":"The technical core of the paper appears sound and the result is a good fit for the journal. The main issue is the unverified applicability of the imported criticality theory to the infinite-range fractional graph. If the authors can show that [37] and [11] apply verbatim to summable infinite-range edge weights, the paper would be in good shape; as written, the proof of Theorem 1.1 contains a genuine structural gap that needs to be closed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I just read Das and de la Fuente-Fernandez's paper on optimal fractional Hardy weights on the discrete half-line. The main theorem is real: it gives an explicit optimal weight W_op_sigma for (-Delta_N)^sigma, the exact threshold alpha=(3+2sigma)/4, and the sharp-at-infinity constant C_sigma for the classical n^{-2sigma} weight, resolving open questions from Gerhat--Krejcirik--Stampach. The extension from the integer line is not routine: the half-line kernel has sign issues, the approximants in Prop 3.2 change sign, and the sigma=1 case needs the Liouville comparison principle instead of a null-sequence. The W_op_1 > W_KPP comparison near infinity is a nice observation, and the unique continuation theorem is a clean corollary.\n\nI checked the central estimates: the Riesz potential identity, the null-sequence decay, the monotonicity of Psi_sigma, and the null-criticality threshold are coherent. The proof of Prop 3.2 using the beta-split and the J_beta growth estimates is sound. No fitted parameters, no circularity.\n\nNow the soft spots. The local finiteness worry: the graph {N, -eK^sigma, R^sigma} has edges of every length, with kernel ~ |m-n|^{-1-2sigma}. But my reading of [37] is that its criticality theory is stated for weighted graphs with finite total edge weight per vertex, exactly what Definition 2.1 gives. Local finiteness was a requirement in the earlier optimal-Hardy paper [36], not obviously in [37]. So the likely fix is a sentence in Section 2 stating that the cited theorems apply to this graph class, not a new proof. A referee should verify this explicitly, but I would bet it goes through. Second, the abstract promises the sharp constant for the classical n^{-2sigma} weight, while the body is careful: C_sigma is the best constant at infinity; the global best constant gamma_H remains open. The abstract overstates. Third, Lemma 3.9 has a sign typo and its Fatou step is terse, but the argument is standard.\n\nThis is a substantive paper, honestly written, with genuine new results. I would bring it to the reading group and would cite it. It deserves peer review; the referee should mainly check the [37] hypothesis question and ask for the clarifying sentence.","headline":"Solid optimal Hardy weight for the discrete half-line; the flagged local-finiteness worry is likely a presentational gap, not a mathematical one.","tokens_in":25141,"tokens_out":9269,"would_cite":true,"duration_ms":94466,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["26D15","26A33"],"pacs":[],"model":"deepseek-v4-flash","headline":"For every σ ∈ (0,1], the paper identifies an explicit weight W_σ^op such that (−∆_N)^σ − W_σ^op is nonnegative, critical, and null-critical on the discrete half-line, and derives the sharp constant for the classical n^{−2σ} Hardy…","keywords":["Hardy inequality","fractional Laplacian","discrete half-line","criticality","optimal Hardy weight","ground state representation","unique continuation"],"falsifier":"For a fixed $\\sigma \\in (0,1]$, evaluate the claimed identity $(-\\Delta_{\\mathbb{N}})^\\sigma I_\\alpha = I_{\\alpha-\\sigma}$ at several $n$ using the explicit kernel in (2.9) and the explicit Gamma formula in (3.2); if the equality fails at any $n$ for $\\alpha = (3+2\\sigma)/4$, the proposed optimal weight is not a Hardy weight.","tokens_in":24066,"feed_emoji":"🧮","tokens_out":19824,"duration_ms":163403,"temperature":0.7,"pith_summary":"The paper proves that the fractional Laplacian $(-\\Delta_{\\mathbb{N}})^\\sigma$ on the discrete half-line $\\mathbb{N}$ has an explicit optimal Hardy weight $W^\\mathrm{op}_\\sigma$ for every exponent $\\sigma \\in (0,1]$. The weight is a ratio of Gamma functions, and subtracting it leaves an operator that is nonnegative, critical, and null-critical, which is the strongest sense in which a Hardy weight can be 'as large as possible'. This answers an open question from prior work about the sharp constant in the Hardy inequality with the classical weight $n^{-2\\sigma}$, giving the upper bound $C_\\sigma = 4^\\sigma \\Gamma((3+2\\sigma)/4)^2 / \\Gamma((3-2\\sigma)/4)^2$. For $\\sigma = 1$ the new weight is pointwise larger than a previously known optimal weight near infinity, so optimal Hardy weights are not unique. As an application, the paper obtains a Landis-type unique continuation result at infinity for fractional Schrödinger equations on $\\mathbb{N}$.","feed_headline":"Explicit optimal Hardy weight found for fractional Laplacian","feed_subtitle":"The weight is null-critical and fixes the sharp constant for n^{-2σ}.","key_machinery":"The load-bearing object is the Riesz potential $I_\\alpha(n)$ of (3.1), defined spectrally as the inner product of Chebyshev basis vectors with $2^{-\\alpha}(1-x)^{-\\alpha}$, and given explicitly in (3.2) as a ratio of Gamma functions with asymptotic $I_\\alpha(n) \\asymp n^{2\\alpha-2}$. Its role is to make the identity $(-\\Delta_{\\mathbb{N}})^\\sigma I_\\alpha = I_{\\alpha-\\sigma}$ (Proposition 3.2) true, which turns the quotient $W_{\\alpha,\\sigma} = I_{\\alpha-\\sigma}/I_\\alpha$ into a Hardy weight via the Agmon–Allegretto–Piepenbrink-type theorem. The ground-state representation (Proposition 3.4) then rewrites the quadratic form of $(-\\Delta_{\\mathbb{N}})^\\sigma - W_{\\alpha,\\sigma}$ as a positive sum of squared differences weighted by $I_\\alpha(n) I_\\alpha(m)$, reducing criticality questions to asymptotics of $I_\\alpha$ and the kernel $K^\\sigma_{m,n}$. The threshold $\\alpha = (3+2\\sigma)/4$ is exactly where $\\sum_n I_\\alpha(n)^2 W_{\\alpha,\\sigma}(n)$ diverges, which is the null-criticality condition that upgrades criticality to optimality.","core_discovery":"On the paper's own terms, the central discovery is that the Riesz potential $I_\\alpha$ defined through Chebyshev polynomials satisfies the transfer identity $(-\\Delta_{\\mathbb{N}})^\\sigma I_\\alpha = I_{\\alpha-\\sigma}$ for $\\sigma < \\alpha < 1+\\sigma$, so the quotient $W_{\\alpha,\\sigma} = I_{\\alpha-\\sigma}/I_\\alpha$ is automatically a Hardy weight. The paper shows that $W_{\\alpha,\\sigma}$ is critical exactly for $\\alpha \\leq (3+2\\sigma)/4$ and null-critical only at the endpoint $\\alpha = (3+2\\sigma)/4$; the endpoint weight $W^\\mathrm{op}_\\sigma(n) = 4^\\sigma \\frac{\\Gamma((3+2\\sigma)/4)^2}{\\Gamma((3-2\\sigma)/4)^2} \\frac{\\Gamma(n-(1+2\\sigma)/4)\\Gamma(n+(5-2\\sigma)/4)}{\\Gamma(n+(-1+2\\sigma)/4)\\Gamma(n+(5+2\\sigma)/4)}$ is therefore optimal. Criticality uses an explicit logarithmic null-sequence for $\\sigma<1$ and the Liouville comparison principle for $\\sigma=1$, while null-criticality follows from the asymptotics $I_\\alpha(n) \\asymp n^{2\\alpha-2}$ and $W_{\\alpha,\\sigma}(n) \\asymp n^{-2\\sigma}$. A corollary is that the best constant at infinity for the classical weight $n^{-2\\sigma}$ equals $C_\\sigma = 4^\\sigma \\Gamma((3+2\\sigma)/4)^2 / \\Gamma((3-2\\sigma)/4)^2$.","pith_inferences":["The same Riesz-potential construction might extend to the subcritical range $\\sigma \\in (1, 3/2)$ if the signed kernel that blocks the graph-Laplacian representation can be handled by a generalized criticality theory; the paper explicitly leaves this as an open question.","Because $W^\\mathrm{op}_\\sigma$ is null-critical, the Rayleigh quotient $\\langle(-\\Delta_{\\mathbb{N}})^\\sigma f, f\\rangle / \\langle f, n^{-2\\sigma} f\\rangle$ over functions supported far out should approach $C_\\sigma$, so the constant could be verified numerically from finite truncations.","The uniqueness (up to scaling) of the Agmon ground state for $(-\\Delta_{\\mathbb{N}})^\\sigma - W^\\mathrm{op}_\\sigma$ suggests a rigidity statement: any positive supersolution of $(-\\Delta_{\\mathbb{N}})^\\sigma$ decaying like a power must be a multiple of $I_{(3+2\\sigma)/4}$.","The contrast with the continuum sharp constant highlights that the lattice changes the constant, not just the formulation; tracing where the difference enters (boundary conditions versus long-range jumps) could transfer the method to other discrete domains."],"forward_implications":["For every $\\sigma \\in (0,1]$, the classical fractional Hardy inequality on $\\mathbb{N}$ with weight $\\gamma n^{-2\\sigma}$ holds for all $0 < \\gamma \\leq C_\\sigma$, and $C_\\sigma = 4^\\sigma \\Gamma((3+2\\sigma)/4)^2 / \\Gamma((3-2\\sigma)/4)^2$ is the best constant at infinity.","The optimal weight $W^\\mathrm{op}_\\sigma$ decays like $n^{-2\\sigma}$ but is null-critical, so no pointwise larger weight can be inserted into the Hardy inequality without destroying it.","For $\\sigma = 1$, $W^\\mathrm{op}_1(n) = \\frac{1}{4}(n^2 - \\frac{9}{16})^{-1}$ exceeds the previously known optimal weight for the standard discrete Laplacian for all sufficiently large $n$, so optimal Hardy weights for a single operator are not unique.","Any solution $u$ of $(-\\Delta_{\\mathbb{N}})^\\sigma u + V u = 0$ with $V \\leq 0$ outside a finite set and $|u| = O(n^{\\sigma-1/2})$ that satisfies $\\liminf_{n\\to\\infty} |u| n^{2-2\\sigma} = 0$ must vanish identically."],"supporting_citations":[{"why":"Supplies the discrete Riesz-potential identity and ground-state representation on Z that Proposition 3.2 extends to the half-line.","marker":"[10]"},{"why":"Provides the spectral representation, kernel decay estimates, the subcriticality range, and the open question about optimal constants that Theorem 1.1 answers.","marker":"[27]"},{"why":"Gives the integer-line optimal Hardy-weight construction whose criticality threshold the paper adapts to N.","marker":"[34]"},{"why":"Supplies the optimal Hardy weight for the standard discrete Laplacian used in the Liouville comparison for σ=1, and the definition of optimality through criticality and null-criticality.","marker":"[36]"},{"why":"Provides the criticality theory for graph Laplacians—null-sequences, ground states, Agmon–Allegretto–Piepenbrink—on which the optimality proof rests.","marker":"[37]"},{"why":"Gives the Liouville comparison principle and the Landis-type unique-continuation theorem used to obtain Theorem 1.2.","marker":"[11]"}],"fun_headline_variants":["Optimal Hardy weight found for discrete fractional Laplacian","Sharp Hardy constant on discrete half-line via optimal weight","Explicit optimal Hardy weight for fractional Laplacian on discrete half-line","Fractional Hardy inequality: optimal weight and sharp constant","Critical Hardy weight optimal for discrete fractional Laplacian"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a theory built for graphs where each point connects to finitely many neighbors still applies to the fractional Laplacian, where every point connects to infinitely many others; the paper uses that theory to prove optimality without an explicit check of its hypotheses.","fun_headline_variants_meta":{"raw":{"variants":["Optimal Hardy weight found for discrete fractional Laplacian","Sharp Hardy constant on discrete half-line via optimal weight","Explicit optimal Hardy weight for fractional Laplacian on discrete half-line","Fractional Hardy inequality: optimal weight and sharp constant","Critical Hardy weight optimal for discrete fractional Laplacian"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001104,"raw_usage":{"total_tokens":4638,"prompt_tokens":1016,"completion_tokens":3622,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":632,"completion_tokens_details":{"reasoning_tokens":3541}},"tokens_in":632,"tokens_out":3622,"duration_ms":28226,"temperature":1.0,"reasoning_tokens":3541,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T18:58:57.634329+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a fixed $\\sigma \\in (0,1]$, evaluate the claimed identity $(-\\Delta_{\\mathbb{N}})^\\sigma I_\\alpha = I_{\\alpha-\\sigma}$ at several $n$ using the explicit kernel in (2.9) and the explicit Gamma formula in (3.2); if the equality fails at any $n$ for $\\alpha = (3+2\\sigma)/4$, the proposed optimal weight is not a Hardy weight.","supporting_citations":[{"cited_title":"Hardy’s inequality for the fractional powers of a discrete Lapla- cian","cited_arxiv_id":null,"evidence_quote":"Supplies the discrete Riesz-potential identity and ground-state representation on Z that Proposition 3.2 extends to the half-line."},{"cited_title":"Criticality transition for positive powers of the discrete Laplacian on the half line.Rev","cited_arxiv_id":null,"evidence_quote":"Provides the spectral representation, kernel decay estimates, the subcriticality range, and the open question about optimal constants that Theorem 1.1 answers."},{"cited_title":"Optimal Hardy inequality for fractional Laplacians on the integers.Ann","cited_arxiv_id":null,"evidence_quote":"Gives the integer-line optimal Hardy-weight construction whose criticality threshold the paper adapts to N."},{"cited_title":"Optimal Hardy inequalities for Schrödinger operators on graphs.Comm","cited_arxiv_id":null,"evidence_quote":"Supplies the optimal Hardy weight for the standard discrete Laplacian used in the Liouville comparison for σ=1, and the definition of optimality through criticality and null-criticality."},{"cited_title":"Criticality theory for Schrödinger operators on graphs.J","cited_arxiv_id":null,"evidence_quote":"Provides the criticality theory for graph Laplacians—null-sequences, ground states, Agmon–Allegretto–Piepenbrink—on which the optimality proof rests."},{"cited_title":"On Landis conjecture for positive Schrödinger operators on graphs.Int","cited_arxiv_id":null,"evidence_quote":"Gives the Liouville comparison principle and the Landis-type unique-continuation theorem used to obtain Theorem 1.2."}],"review_version":1}