REVIEW
Spectral convergence associated with Weil's quadratic form
Number Theory · Submitted 23 Oct 2026 · 11 references
Pith's verdict: Real advance in the Weil-form spectral program: monotone upper approximation of zeta ordinates and an entire-function limit, with a coherent compression proof; the advertised 'no zeta-zero input' claim, however, rests on an equivalence that is asserted but never displayed.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
The $n$-th eigenvalue of a finite-interval problem built purely from prime powers stays above $\gamma_n$ and descends monotonically onto it as the interval grows, and the normalized spectral functions converge to the canonical zero product.
The reading
What carries the argument
The load-bearing object is the inverse-compression identity $(D_{a,\pi})^{-1} = j_a^* D_{\pi/2}^{-1} j_a$ (Theorem 3.3): the inverse of the distinguished finite-interval extension equals the compression, through the isometric embedding $j_a: H(A_a) \to H_W$, of the inverse of the global operator $D_{\pi/2}$ whose spectrum is the set $\Gamma$ of distinct zero ordinates. Around this identity the argument uses four supporting structures: (i) the Weil quadratic form $Q_W(u,f) = \sum_{\gamma\in\Gamma} m_\gamma \hat{u}(\gamma)\overline{\hat{f}(\gamma)}$, whose restriction to $[-a,a]$ defines the operator $A_a$ and, via the coercivity estimate $Q_W(u) \ge \lambda_a \|u\|^2$ with $\lambda_a > 0$, the Hilbert space $H(A_a)$; (ii) von Neumann's theory of self-adjoint extensions of the densely defined symmetric operator $i\,d/dx$, with the phase $\theta = \pi$ singled out by reflection symmetry and by the absence of a zero eigenvalue; (iii) the spectral function $W(a,\pi;z) = (z-i)J_a(-i,z) - (z+i)J_a(i,z)$ built from reproducing kernels of the de Branges space $\mathcal{F}H(A_a)$, whose zeros are exactly the eigenvalues of $D_{a,\pi}$; and (iv) the global realization of $D_{\pi/2}$ on the model space generated by $E(z) = \xi(1/2-iz) + \xi'(1/2-iz)$. The convergence argument runs on two rails: the Hilbert-Schmidt approximation $\Pi_a C \Pi_a \to C$ with $C = D_{\pi/2}^{-1}$ and $\Pi_a \to I$ strongly, feeding the max-min principle for compact operators, and the Hadamard factorization of $W(a,\pi;\cdot)$ into $\prod_n (1 - z^2/t_{a,n}^2)$, whose tails are controlled uniformly in $a$ by the inequality $t_{a,n} \ge \gamma_n$.
Extended reading notes
Core claim
The paper's central claim is that the distinct positive ordinates $\Gamma^+ = \{\gamma_1 < \gamma_2 < \cdots\}$ of the nontrivial zeros of the Riemann zeta function are recovered as the $a \to \infty$ limit of the positive spectra of a distinguished family of self-adjoint operators, where each finite-stage operator is defined purely by local arithmetic data (the archimedean place together with the finitely many prime powers $p^m \le e^{2a}$), with no zero information as input. Under RH, Theorem 1.1 states that for the extension $D_{a,\pi}$ of the differentiation operator $i\,d/dx$ on the Hilbert space $H(A_a)$ determined by the Weil quadratic form localized to $[-a,a]$, the $n$-th positive eigenvalue satisfies $t_{a,n} \ge \gamma_n$ and decreases monotonically to $\gamma_n$ as $a \to \infty$. Theorem 1.2 upgrades this to the level of entire functions: the normalized spectral functions $W(a,\pi;z)/W(a,\pi;0)$ converge locally uniformly on $\mathbb{C}$ to the canonical product $\Delta_\Gamma(z) = \prod_{\gamma \in \Gamma^+} (1 - z^2/\gamma^2)$ over the distinct ordinates, which does not record multiplicities. The engine of the proof is Theorem 3.3, the inverse-compression identity $(D_{a,\pi})^{-1} = j_a^* D_{\pi/2}^{-1} j_a$: the inverse of the local extension is the compression to $H(A_a)$ of the inverse of a global self-adjoint operator $D_{\pi/2}$ whose spectrum is exactly $\Gamma$. Compressing this Hilbert-Schmidt inverse converges in Hilbert-Schmidt norm to the global inverse, the max-min principle turns that convergence into the monotone eigenvalue bounds, and Hadamard factorization of the even entire function $W(a,\pi;\cdot)$ with the tail control supplied by $t_{a,n} \ge \gamma_n$ yields the locally uniform entire-function limit.
Load-bearing premise
The argument stands or falls on the imported fact that the localized Weil quadratic form on each finite interval has a strictly positive spectral gap, $Q_W(u) \ge \lambda_a \|u\|^2$ with $\lambda_a > 0$, and is genuinely determined by the finitely many prime powers $p^m \le e^{2a}$ together with the archimedean place, since without both properties the Hilbert space $H(A_a)$ is not a true subspace of $L^2(-a,a)$, the deficiency-index computation collapses, and the inverse-compression identity (3.5) cannot stand.
Editorial extensions
If this is right
- For every fixed $n$, the bound $t_{a,n} \ge \gamma_n$ yields a computable, monotone family of upper bounds on the $n$-th distinct zero ordinate: each finite-interval spectral problem (with data only the prime powers $p^m \le e^{2a}$) traps $\gamma_n$ from above at every finite $a$ (Theorem 1.1).
- The locally uniform limit $W(a,\pi;z)/W(a,\pi;0) \to \Delta_\Gamma(z)$ packages the whole array of distinct ordinates into one entire function, and Corollary 5.1 converts this into the RH-equivalent limit formula of the companion paper, with the $a$-dependence entering only through a nonzero scalar.
- If all zeta zeros are simple, the normalized spectral function converges to $\xi(1/2 - iz)/\xi(1/2)$ on $\mathbb{C}$, so the Riemann xi function itself is recovered as the limit of arithmetically defined spectral functions (Corollary 5.2).
- The resolvents converge as well: transported to the global space, $(D_{a,\pi} - z)^{-1}$ converges to $(D_{\pi/2} - z)^{-1}$ in operator norm, uniformly on compact sets avoiding $\Gamma$ (Remark 4.1).
- Multiplicities are invisible to the limit: because $\Delta_\Gamma$ runs over distinct ordinates, the construction cannot detect repeated zeros, and neither theorem requires or uses the simplicity of the zeta zeros.
Reading between the lines
- Equivalence reading (editorial): since the finite-interval problems need no zero data, and since the companion paper's Corollary 1.6 asserts that a limit formula of the proved shape implies RH, the paper in effect converts RH into an unconditional convergence statement about explicitly computable Gram-matrix spectral problems; the rate of that convergence is not treated here and is the natural nex
- The numerics suggest a steep practical tradeoff worth probing: moving $X$ from 3 to 11 shrinks the sixth-zero error by roughly fourteen orders of magnitude while the Gram-matrix condition number inflates from $10^7$ to $10^{34}$, so conditioning, not arithmetic data, is the method's ceiling; a better-conditioned basis could push the scheme further and, by contraposition, any rigorously verified ne
- The nested projective picture (editorial): the monotone descent onto $\Gamma$ is a consequence of the subspaces $M_a = j_a H(A_a)$ filling $H_W$, which suggests studying the family as a limit of de Branges spaces taken in a resolvent sense; making that limit precise as a statement about reproducing kernels could clarify why only the distinct ordinates, and no multiplicity data, can survive.
Original abstract
Assuming the Riemann hypothesis, we study a distinguished self-adjoint extension of the differentiation operator on a Hilbert space defined by the Weil quadratic form localized to a finite interval. The local quadratic form is determined by data from finitely many places, without requiring the zeros of the Riemann zeta function as input. We prove that the positive eigenvalues of this extension converge to the distinct positive ordinates of the nontrivial zeros of the Riemann zeta function as the interval expands. We also prove that the associated normalized spectral entire functions converge locally uniformly on the complex plane to the canonical product over these distinct ordinates. The key step is an identity expressing the inverse of the local extension as a compression of the inverse of a global self-adjoint operator.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Assuming the Riemann hypothesis, we study a distinguished self-adjoint extension of the differentiation operator on a Hilbert space defined by the Weil quadratic form localized to a finite interval. The local quadratic form is determined by data from finitely many places, without requiring the zeros of the Riemann zeta function as input. We prove that the positive eigenvalues of this extension converge to the distinct positive ordinates of the nontrivial zeros of the Riemann zeta function as the interval expands. We also prove that the associated normalized spectral entire functions converge locally uniformly on the complex plane to the canonical product over these distinct ordinates. The key step is an identity expressing the inverse of the local extension as a compression of the inverse of a global self-adjoint operator.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Assuming the Riemann hypothesis, we study a distinguished self-adjoint extension of the differentiation operator on a Hilbert space defined by the Weil quadratic form localized to a finite interval. | contribution 0.97 | 0.85 | 0.07 | asserted 0.83 |
| s1 | The local quadratic form is determined by data from finitely many places, without requiring the zeros of the Riemann zeta function as input. | method 0.83 | 0.68 | 0.04 | asserted 1.00 |
| s2 | We prove that the positive eigenvalues of this extension converge to the distinct positive ordinates of the nontrivial zeros of the Riemann zeta function as the interval expands. | result 1.00 | 0.95 | 0.78 | asserted 1.00 |
| s3 | We also prove that the associated normalized spectral entire functions converge locally uniformly on the complex plane to the canonical product over these distinct ordinates. | result 1.00 | 0.94 | 0.16 | asserted 1.00 |
| s4 | The key step is an identity expressing the inverse of the local extension as a compression of the inverse of a global self-adjoint operator. | method 0.97 | 0.86 | 0.05 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.