{"id":"f5f0c4f2-9055-4d3e-a7d8-607bab403021","arxiv_id":"2506.16178","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For Coulombic many-electron eigenfunctions, the occupation-number eigenvalues decay at least as k^{-8/3} (density matrix) and k^{-2} (kinetic energy matrix), improving to k^{-10/3} and k^{-8/3} when the wavefunction vanishes at electron-electron coalescence.","lead":"This paper proves rigorous decay estimates for the eigenvalues of the one-electron density matrix and kinetic energy density matrix of atomic eigenfunctions. It shows faster decay when the wavefunction vanishes at electron-electron collision points, which covers fermionic atoms.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorems 1.1–1.2 are well-posed only for real-valued eigenfunctions: kernels (1.3)–(1.4) omit the complex conjugate, so Int(γ) and Int(τ) are not self-adjoint and ρ in (1.7) is not the diagonal of γ for complex ψ.","rationale":"I read the paper's central claim as explicit decay bounds for eigenvalues of integral operators associated with a Coulombic eigenfunction. The proof's engine is the self-adjoint factorization (1.13). The definitions (1.3)–(1.4) and the norm (1.7) are mutually consistent only for real ψ. This is an internal inconsistency, not a disagreement with consensus, and it affects the statement of both theorems. The hydrogenic example shows the failure is real rather than cosmetic. The reader's weakest_assumption identified exactly this issue; I agree. I do not see an equally load-bearing problem elsewhere: the regularity estimates from [15] and [20] are cited appropriately, the Fourier argument in Lemma 4.3 is coherent, and Theorem 4.1's Schatten-class conclusion matches the claimed exponents. Novelty overlap with [20] is a literature-assessment matter, not a correctness risk. Therefore the reader's CONDITIONAL verdict is appropriate; my stress-test does not change it.","tokens_in":18745,"tokens_out":7808,"duration_ms":87077,"concrete_test":"Take N=1 and ψ=ψ_{nlm} with m≠0, a complex eigenfunction of the hydrogenic Hamiltonian H=-Δ-Z/|x|. Compute the operator T with kernel ψ(x)ψ(y) from (1.3) and the operator Ψ*Ψ from (1.13). Show that T is not self-adjoint, that its nonzero eigenvalue is ∫ψ² (generally complex), and that the kernel of Ψ*Ψ is \\overline{ψ(y)}ψ(x), which differs from ψ(x)ψ(y). This settles that Theorems 1.1–1.2 fail as stated for complex eigenfunctions and that the missing real-valued assumption or the missing complex conjugate in (1.3)–(1.4) is load-bearing.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing point is a missing hypothesis. The factorization (1.13) gives Int(γ)=Ψ*Ψ, whose kernel is ∫ \\overline{ψ(x̂,y)}ψ(x̂,x)dx̂, whereas definition (1.3) is ∫ ψ(x̂,x)ψ(x̂,y)dx̂. These coincide only if ψ is real. The same mismatch occurs for Int(τ) and V*V. Moreover, (1.7) defines ρ(x)=γ(x,x)=∫|ψ(x̂,x)|²dx̂, but inserting (1.3) on the diagonal gives ∫ψ(x̂,x)²dx̂, so the right-hand side norms in Theorems 1.1–1.2 do not control the stated operators unless ψ is real. A concrete failure: take N=1 and a hydrogenic eigenfunction ψ_{nlm} with m≠0. Then Int(γ) is the rank-one operator f↦ψ∫ψ(y)f(y)dy in L²(R³); its only nonzero eigenvalue is ∫ψ², generally non-real, so λ_k(Int(γ)) as positive eigenvalues do not exist. Since H is real, each eigenspace has a real basis, and the proof is valid after adding the assumption that ψ is real or after inserting conjugates in (1.3)–(1.4); the defect is a stated-hypothesis/notation gap rather than a flaw in the estimates. Nevertheless, as written the central claim is undefined for a natural class of eigenfunctions.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies the one-particle density matrix γ and the kinetic energy density matrix τ associated with an eigenfunction ψ of the N-electron atomic Schrödinger operator H = -Δ + V. The main results are the eigenvalue bounds λ_k(Int(γ)) ≤ C k^{-8/3} ρ_{3/8} and λ_k(Int(τ)) ≤ C k^{-2} ρ_{1/2} (Theorem 1.1), and improved bounds λ_k(Int(γ)) ≤ C k^{-10/3} ρ_{3/10} and λ_k(Int(τ)) ≤ C k^{-8/3} ρ_{3/8} when ψ vanishes at electron-electron coalescence points (Theorem 1.2). The constants depend only on E and N, and the right-hand sides contain explicit norms of the one-particle density ρ. The proof factorizes Int(γ) and Int(τ) as Ψ*Ψ and V*V, estimates the singular values of Ψ and V by bounding Fourier coefficients of the kernels via derivative estimates (2.3) and (2.5), and then uses Schatten/quasi-norm machinery. The exposition is clear and the proof strategy is direct and essentially self-contained, conditional on the regularity results from Fournais–Sørensen.","tokens_in":18933,"tokens_out":20546,"duration_ms":210267,"significance":"The proof strategy is transparent and, conditional on the cited regularity estimates, avoids the general Birman–Solomyak apparatus by directly estimating Fourier coefficients of the kernels. The explicit dependence of the bounds on the eigenfunction through lattice norms of ρ is a genuine improvement over previous results with implicit constants, and the improved decay rates under the antisymmetry/vanishing condition appear to be new. No parameters are fitted, and the earlier asymptotic results are used only to indicate sharpness, not inside the derivation. The main obstacle to accepting the paper in its present form is the incomplete definition of the density matrices for complex-valued eigenfunctions, which makes the central statements ill-posed as written; this is a fixable but load-bearing gap.","major_comments":[{"comment":"The kernels γ0 and τ0 are defined without complex conjugation. For a complex-valued eigenfunction ψ, the operator Int(γ0) with kernel γ0(x,y)=∫ψ(ˆx,x)ψ(ˆx,y)dˆx is not self-adjoint in general, so the eigenvalues λ_k(Int(γ0)) are not defined. The factorization (1.13) gives Int(γ)=Ψ*Ψ, whose kernel is ∫ \\overline{ψ(ˆx,y)}ψ(ˆx,x)dˆx, not the kernel in (1.3) unless ψ is real. Similarly, ρ(x)=γ(x,x) in (1.7) equals ∫|ψ|^2 only after conjugation. A concrete failure occurs for N=1 and a hydrogenic eigenfunction with m≠0: Int(γ) is rank-one with sole eigenvalue ∫ψ^2, generally non-real. Since H is real, the results can be salvaged either by assuming ψ is real, or, preferably, by inserting the complex conjugate in (1.3)–(1.4) (and in the definition of τ). The proofs in Section 5 go through unchanged because the singular-value estimates apply identically to kernels with \\overline{ψ}.","section":"Section 1, Eqs. (1.3)–(1.4) and Theorems 1.1–1.2"},{"comment":"The kernel b(ˆx)ψ(ˆx,x) is said to satisfy condition (4.1) with α=1 and z_j(x)=x_j, j=1,...,N-1. The regularity bound (2.3) used here involves λ(x)=min{|x|, 1/√2 |x-x_k|}, which includes the nuclear singularity at x=0. Thus the point x=0 must be included among the z_k in (4.1); as written, the stated list of N-1 points does not cover it. This is evidently a typographical omission—Lemma 4.3 and Theorem 4.1 allow N arbitrary singular points—but it should be corrected in the proof.","section":"Section 5.1, proof of Theorem 5.1"}],"minor_comments":[{"comment":"The finiteness of the norms ρ_{3/8}, ρ_{1/2}, and ρ_{3/10} is not stated as a hypothesis. The bounds are only meaningful when these norms are finite, as noted in the text after Theorem 1.1; it would be clearer to include this assumption in the theorem statements.","section":"Theorems 1.1 and 1.2"},{"comment":"The phrase 'a.e. x ∈ R3, a.e. y ∈ R3' would be cleaner as 'a.e. x, y ∈ R3'.","section":"Section 1, after (1.3)"},{"comment":"The summation over 'p,s:|p|+|s|=κ' should state explicitly that p and s are multi-indices and that the sum runs over all such multi-indices.","section":"Proof of Lemma 4.3"},{"comment":"The notation '0≤k<m' in the summation is non-standard; it should be clarified that the sum runs over multi-indices k with k_j ≤ m_j for all j and k ≠ m.","section":"Eq. (2.23)"}],"recommendation":"major_revision","confidential_remarks":"The complex-conjugate gap in the definitions is the main issue and should be fixed before publication, either by adding a real-valuedness assumption or, preferably, by inserting the missing conjugates. The missing z=0 in the proof of Theorem 5.1 is a typo but needs correction. Aside from these points, the mathematical core appears sound and the paper is a solid contribution to the spectral theory of reduced density matrices."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper proves eigenvalue decay bounds for the one-particle density and kinetic energy density matrices of Coulombic eigenfunctions. Theorem 1.1 gives explicit bounds of order k^{-8/3} and k^{-2} with constants depending only on E, N, and the one-particle density; Theorem 1.2 improves these to k^{-10/3} and k^{-8/3} under the condition that the wavefunction vanishes at electron–electron coalescence points, which holds for antisymmetric states. The proofs are direct and self-contained: the singular values of the integral operators are controlled via Fourier decay of the kernels, fed by the regularity estimates of Fournais–Sørensen and the enhanced regularity from Hearnshaw.\n\nThe main soft spot is a definitional gap. The kernels (1.3) and (1.4) omit the complex conjugate, so for complex-valued eigenfunctions the operators Int(γ) and Int(τ) are not self-adjoint and their eigenvalues are not defined. The factorization (1.13), which is what the proofs actually use, defines the self-adjoint operators with the conjugate in the kernel. Thus the theorems as stated are only well-posed for real ψ. This is easily fixed either by assuming ψ is real (every eigenspace of this real operator has a real basis) or by inserting the conjugate in the definitions. It is a notation/hypothesis gap, not a flaw in the estimates.\n\nA second issue: the claim that Theorem 1.2 is new depends on what exactly is in [20]. The author says only that it contains 'some results in this direction'. A referee should ask for a precise comparison. Also, the theorems would be cleaner if the right-hand sides were explicitly stated to be finite (the author notes this in passing).\n\nThe math is solid. The singular value machinery is standard and the regularity arguments are coherent. The paper is written for specialists in spectral theory and the theory of reduced density matrices; it should be useful to anyone studying finite-dimensional approximation errors in quantum chemistry.\n\nIt deserves serious peer review. I would accept it after the authors fix the complex conjugate issue and clarify the relationship to Hearnshaw's paper.","headline":"Solid eigenvalue bounds for Coulombic density matrices; fix the missing complex conjugate and clarify the overlap with Hearnshaw.","tokens_in":19579,"tokens_out":9741,"would_cite":true,"duration_ms":97418,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35J10","47G10","81Q10"],"pacs":[],"model":"deepseek-v4-flash","headline":"For an atomic bound state, the eigenvalues of the one-particle density matrix decay at least as $k^{-8/3}$, and for antisymmetric states at least as $k^{-10/3}$.","keywords":["multi-particle Schrödinger operator","one-particle density matrix","kinetic energy density matrix","occupation numbers","eigenvalue estimates","integral operators","singular values","Coulombic wave functions"],"falsifier":"Multiply a real bound state $\\phi$ by the global phase $e^{i\\pi/4}$; with the paper's definition (1.3), $\\mathrm{Int}(\\gamma)$ has kernel $i\\,\\phi(\\hat{x},x)\\phi(\\hat{x},y)$, which is not self-adjoint, so the theorem's hypothesis that the operator is self-adjoint and non-negative fails.","tokens_in":18412,"feed_emoji":"⚛️","tokens_out":8994,"duration_ms":92640,"temperature":0.7,"pith_summary":"The paper proves quantitative decay bounds for the eigenvalues of the one-particle density matrix and the kinetic energy density matrix built from a bound-state eigenfunction of an atom with $N$ electrons. The main bounds are $\\lambda_k(\\mathrm{Int}(\\gamma)) \\le C k^{-8/3}$ and $\\lambda_k(\\mathrm{Int}(\\tau)) \\le C k^{-2}$, with constants that depend only on the energy and the particle number, not on the eigenfunction, and with the eigenfunction entering only through an explicit norm of the one-particle density. When the eigenfunction vanishes at particle coalescence points, which is true for totally antisymmetric (fermionic) wavefunctions, the bounds improve to $k^{-10/3}$ and $k^{-8/3}$. These exponents match previously established eigenvalue asymptotics, so the bounds are sharp in their decay order.","feed_headline":"Fermionic atomic states yield faster occupation-number decay","feed_subtitle":"New eigenvalue bounds: fermionic states have occupation numbers decaying at least as k^(-10/3), beating the universal k^(-8/3).","key_machinery":"The carrying mechanism is a reduction of the eigenvalue problem to singular-value estimates for integral operators. The factorizations $\\mathrm{Int}(\\gamma) = \\Psi^*\\Psi$ and $\\mathrm{Int}(\\tau) = V^*V$ tie the eigenvalues to squares of singular values of $\\Psi$ and $V$, whose kernels are $\\psi$ and $\\nabla\\psi$. The main technical tool is a new singular-value bound (Theorem 4.1) for integral operators whose kernels satisfy derivative bounds with explicit factors depending on the distance to the coalescence set; the proof uses Fourier decay of such kernels (Lemma 4.3) and a sum over lattice cubes of weighted norms. Regularity input comes from derivative estimates for Coulombic eigenfunctions, with an improved version under the vanishing condition (1.10) that yields the faster decay.","core_discovery":"On its own terms, the paper establishes that the singular values of the integral operators with kernels $\\psi(\\hat{x},x)$ and $\\nabla_x \\psi(\\hat{x},x)$ obey power-law decay with exponents determined by the regularity of $\\psi$ at the coalescence points of the Coulomb potential. Concretely, $\\lambda_k(\\mathrm{Int}(\\gamma)) \\le C k^{-8/3} \\rho_{3/8}$ and $\\lambda_k(\\mathrm{Int}(\\tau)) \\le C k^{-2} \\rho_{1/2}$ always hold, and the vanishing condition (1.10) improves these to $\\lambda_k(\\mathrm{Int}(\\gamma)) \\le C k^{-10/3} \\rho_{3/10}$ and $\\lambda_k(\\mathrm{Int}(\\tau)) \\le C k^{-8/3} \\rho_{3/8}$. The constants do not depend on $\\psi$, but may depend on $E$ and $N$; the norms $\\rho_q$ are sums over unit cubes of powers of the one-particle density. The proof requires only polynomial decay of $\\psi$ at infinity, not the exponential bound assumed in earlier work.","pith_inferences":["The faster decay for antisymmetric states suggests that Pauli exclusion itself drives the improved summability of occupation numbers, which could be tested numerically by comparing singlet and triplet pair configurations in few-electron atoms.","The same Fourier-decay machinery likely yields analogous bounds for $n$-particle reduced density matrices with $n>1$, where the coalescence sets have more complex geometry.","If the implicit real-valued assumption were relaxed, the density matrix would need a complex conjugate in its kernel; the stated results apply to real eigenfunctions, which always exist for this real Hamiltonian.","One could probe sharpness by constructing model eigenfunctions with cusp-like singularities producing Fourier decay exactly at the boundary exponents."],"forward_implications":["Occupation numbers of an atomic bound state are guaranteed to decay at least as $k^{-8/3}$, and at least as $k^{-10/3}$ for antisymmetric states.","The eigenvalues of the kinetic energy density matrix decay at least as $k^{-2}$, improving to $k^{-8/3}$ under the vanishing condition.","Because the constants do not depend on the eigenfunction, the bounds apply uniformly across all bound states of a given atom with fixed energy and electron number.","The statements extend to molecules with fixed nuclei, as noted in the paper.","These bounds can be used to control errors in finite-basis approximations used in quantum chemistry."],"supporting_citations":[{"why":"Supplies bounds on derivatives of the eigenfunction of all orders with explicit distance-to-coalescence dependence, used for the Fourier decay in Theorem 4.1.","marker":"[15]"},{"why":"Provides the sharp regularity at coalescence points from which the improved second-derivative bounds under condition (1.10) are derived.","marker":"[13]"},{"why":"Gives the refined method (Proposition 2.6) used to upgrade the derivative bounds to the vanishing-case rates.","marker":"[21]"},{"why":"Earlier eigenvalue bound for Int($\\gamma$) under exponential decay; the present paper improves it by making constants explicit.","marker":"[34]"},{"why":"Earlier eigenvalue bound for Int($\\tau$) and the asymptotic rate $k^{-2}$ that the new bound matches.","marker":"[33]"},{"why":"Establishes the eigenvalue asymptotics $k^{-8/3}$ for Int($\\gamma$), confirming the sharpness of the new exponent.","marker":"[32]"},{"why":"Provides the quasi-norm machinery, in particular the inequality used to sum over lattice cubes (Proposition 3.2).","marker":"[2]"},{"why":"States the boundedness of second $x$-derivatives of the transformed eigenfunction under the vanishing condition, leading to Theorem 2.2.","marker":"[20]"}],"fun_headline_variants":["Fermionic atomic states yield faster occupation-number decay","Antisymmetric states boost decay of atomic density-matrix eigenvalues","New bounds: Coulombic density matrices decay at k^(-10/3) for fermions","Explicit eigenfunction-based eigenvalue bounds for atomic matrices"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proofs assume the eigenfunction is real-valued; for complex eigenfunctions the density kernel (1.3) is not Hermitian, so the stated eigenvalues $\\lambda_k(\\mathrm{Int}(\\gamma))$ are not defined.","fun_headline_variants_meta":{"raw":{"variants":["Fermionic atomic states yield faster occupation-number decay","Antisymmetric states boost decay of atomic density-matrix eigenvalues","New bounds: Coulombic density matrices decay at k^(-10/3) for fermions","Explicit eigenfunction-based eigenvalue bounds for atomic matrices"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000308,"raw_usage":{"total_tokens":1842,"prompt_tokens":1107,"completion_tokens":735,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":723,"completion_tokens_details":{"reasoning_tokens":662}},"tokens_in":723,"tokens_out":735,"duration_ms":8351,"temperature":1.0,"reasoning_tokens":662,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T23:46:43.988565+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Multiply a real bound state $\\phi$ by the global phase $e^{i\\pi/4}$; with the paper's definition (1.3), $\\mathrm{Int}(\\gamma)$ has kernel $i\\,\\phi(\\hat{x},x)\\phi(\\hat{x},y)$, which is not self-adjoint, so the theorem's hypothesis that the operator is self-adjoint and non-negative fails.","supporting_citations":[{"cited_title":"Fournais and T","cited_arxiv_id":null,"evidence_quote":"Supplies bounds on derivatives of the eigenfunction of all orders with explicit distance-to-coalescence dependence, used for the Fourier decay in Theorem 4.1."},{"cited_title":"Fournais, M","cited_arxiv_id":null,"evidence_quote":"Provides the sharp regularity at coalescence points from which the improved second-derivative bounds under condition (1.10) are derived."},{"cited_title":"Hearnshaw and A","cited_arxiv_id":null,"evidence_quote":"Gives the refined method (Proposition 2.6) used to upgrade the derivative bounds to the vanishing-case rates."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier eigenvalue bound for Int($\\gamma$) under exponential decay; the present paper improves it by making constants explicit."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Earlier eigenvalue bound for Int($\\tau$) and the asymptotic rate $k^{-2}$ that the new bound matches."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the eigenvalue asymptotics $k^{-8/3}$ for Int($\\gamma$), confirming the sharpness of the new exponent."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the quasi-norm machinery, in particular the inequality used to sum over lattice cubes (Proposition 3.2)."},{"cited_title":"Eigenvalue Bounds for Multi-Particle Reduced Density Matrices of Coulombic Wavefunctions","cited_arxiv_id":"2412.16073","evidence_quote":"States the boundedness of second $x$-derivatives of the transformed eigenfunction under the vanishing condition, leading to Theorem 2.2."}],"review_version":1}