Salvato in:
Dettagli Bibliografici
Autore principale: Thomson, David
Natura: Recurso digital
Lingua:inglese
Pubblicazione: Zenodo 2026
Soggetti:
Accesso online:https://doi.org/10.5281/zenodo.18190604
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
Sommario:
  • <p>Operator Package 5 advances the Operator Package series by converting the finite-window, coherence-filtered operator mechanics from a <em>scanner</em> into a <em>generator</em>. In Packages 2--3, finite log-scale windows and holonomy-compatible boundary closure induce a discrete mode grid, and a compact APM-derived kernel operator acts as a coherence filter on that grid. A proof-oriented RH program, however, requires a characteristic object whose vanishing condition targets the zero set. Package 5 supplies this by introducing a one-parameter <em>log-phase twist</em> and an associated determinant closure.</p> <p>On a finite log-scale window $I=[q_-,q_+]$ of length $L_q$, define the windowed convolution operator<br>\[<br>(W_I f)(q) = \int_{q_-}^{q_+}\Phi_{\mathrm{APM}}(q-u)\,f(u)\,du,<br>\]<br>and the unitary twist<br>\[<br>(U_t f)(q)=e^{itq}f(q).<br>\]<br>The twisted kernel family is defined by unitary conjugation,<br>\[<br>W_I(t)=U_t\,W_I\,U_t^{-1},<br>\]<br>equivalently giving a twisted kernel factor $e^{it(q-u)}$ on the finite window. With a holonomy-compatible self-adjoint Hamiltonian $H_{\mathrm{APM}}$ and energy cutoff $E_c$, define the coherence projector<br>\[<br>\Pi_{E_c}=\mathbf{1}_{(-\infty,E_c]}(H_{\mathrm{APM}}),<br>\]<br>and the coherence-selected finite-rank family<br>\[<br>S_{E_c}(t)=\Pi_{E_c}\,W_I(t)\,\Pi_{E_c}.<br>\]<br>The Package 5 generator is the finite-rank determinant<br>\[<br>D_{E_c}(t)=\det\!\big(S_{E_c}(t)\big),<br>\]<br>taken on the finite-dimensional range of $\Pi_{E_c}$.</p> <p>Package 5 proves that $D_{E_c}(t)$ extends to an even entire function of the twist parameter $t$ and establishes an exact diagonalization in the calibration regime $V_{\mathrm{APM}}\equiv 0$, where the holonomy Fourier basis simultaneously diagonalizes the twisted family. In that regime, each diagonal eigenvalue becomes an explicit Fejer-type smoothing of the APM multiplier $\Xi_{\mathrm{APM}}$ centered at a shifted frequency, and the determinant becomes a finite product of these shifted diagonal samples. The first closed limit theorem in this package is the large-window limit at fixed finite rank: as $L_q\to\infty$, the Fejer-smoothed diagonal samples converge pointwise (and locally uniformly) to $\Xi_{\mathrm{APM}}$, yielding a finite product limit for $D_{E_c}(t)$ in terms of shifted $\Xi_{\mathrm{APM}}$ factors.</p> <p>To address the cutoff-removal step needed for a canonical infinite-dimensional closure, Package 5 then introduces a standard Sobolev-weighted trace-class renormalization based on the log-scale Hamiltonian. For $s>\tfrac12$, set<br>\[<br>A_s=(I+H_{\mathrm{APM}})^{-s/2},\qquad K_s(t)=A_s\,W_I(t)\,A_s,<br>\]<br>which is trace class on a finite window, and define the Fredholm determinant generator<br>\[<br>\Delta_s(t)=\det\!\big(I+\lambda K_s(t)\big).<br>\]<br>Package 5 proves analyticity (entire extension) and evenness for $\Delta_s(t)$ under the same APM kernel symmetry assumptions, and isolates the remaining RH-critical bottleneck as a precise closure target: after a unique APM normalization and window removal, the limiting determinant generator should identify with $\Xi_{\mathrm{APM}}(t)$. With the Package 4 identification $\Xi_{\mathrm{APM}}\equiv \Xi$, this provides a direct Hilbert--Polya-class generator route in which the Riemann $\Xi$ arises as an operator determinant of a self-adjoint compact family, leaving the final fork in the positivity / canonical-system domain (the mechanism that forces determinant zeros onto the real $t$-axis).</p> <p>This record is the determinant module of the RH program: it formalizes the twisted-kernel generator, proves the finite-rank analytic properties and the calibration diagonalization, closes the large-window limit at fixed cutoff, and provides the canonical trace-class framework required for the cutoff-removal determinant program.</p>