Overview
The Foundation paper establishes that the admissible fibre \(F_n\) carries the Weil representation \(V_\rho\) of \(\mathrm{Heis}_3(\mathbb{Z}/q\mathbb{Z})\) on \(L^2(\mathbb{Z}/q\mathbb{Z})\). Q5a addresses the next question: how this discrete structure approaches a continuum operator theory as \(q\to\infty\).
The paper deliberately avoids any set-level identification \(\mathbb{Z}/q\mathbb{Z}\to\mathbb{R}\). Instead, it formulates the limit in the category of Hilbert spaces and operator algebras, where the relevant objects are embeddings, forms, generators, and resolvents.
The central output is a conditional Mosco convergence theorem for the admissibility forms \(\mathcal{E}_q\), with limiting operator \(L_\Pi=-A\partial_x^2\) on \(L^2(\mathbb{R})\).
Q5a also removes a potential structural obstruction: the full Weil-block embedding hypothesis \([H1]\) is not needed in full generality, because admissibility confines the relevant profiles to a low-frequency sector where the sinc embedding is canonical.
Main contributions
- Categorical clarification: the continuum limit is formulated in \(\mathrm{Hilb}\) and \(C^*\)-algebras, not as a pointwise map from a finite set to \(\mathbb{R}\).
- Hilbert limit: \(C_q=L^2(\mathbb{Z}/q\mathbb{Z})\) embeds into \(L^2(\mathbb{R})\) through the sinc/WSK construction, giving the correct additive Hilbert limit.
- Low-frequency resolution of \([H1]\): full Weil-block compatibility remains a representation-theoretic refinement, but it is not required in the admissible sector relevant to \(T1\)-\(T3\).
- Representational convergence: the rescaled Weil generators \(\widehat{X}_q,\widehat{P}_q\) are stated to converge to the metaplectic generators \(x\) and \(-i\partial_x\), conditional on \([H2]\).
- Mosco programme: the admissibility forms \(\mathcal{E}_q\) converge conditionally to the Dirichlet form \(\mathcal{E}(f,g)=A\int_{\mathbb{R}}\partial_xf\,\partial_xg\,dx\).
- Empirical tightness: the spectral tail test gives \(\mathcal{E}_q(q/3)/\|f_q\|^2<2.1\times 10^{-5}\) for all tested primes \(q\in\{29,61,101,151\}\).
The three convergence statements
Q5a decomposes the large-\(q\) problem into three logically separated convergence statements.
- \(T1\): Hilbert inductive limit. The spaces \(C_q\) form a directed Hilbert system whose continuum completion is \(L^2(\mathbb{R})\), with \(\mathcal{S}(\mathbb{R})\) as dense test space.
- \(T2\): representational convergence. The rescaled finite Weil generators converge, on the test domain, to the standard metaplectic generators of \(\mathrm{Heis}_3(\mathbb{R})\).
- \(T3\): Mosco convergence. The rescaled admissibility forms \(\mathcal{E}_q\) converge to a closed Dirichlet form on \(L^2(\mathbb{R})\), whose self-adjoint operator is \(L_\Pi=-A\partial_x^2\).
Status of the hypotheses
A central purpose of Q5a is to separate proved structure, conditional input, numerical support, and open analytic work.
| Label | Content | Status in Q5a |
|---|---|---|
| \([H\text{-}\Pi1]\) | \(\Pi_q\) orthogonal and parity-compatible | Proved in the O-series |
| \([H\text{-}\Pi2]\) | Shell locality, \(n_*(q)\to\infty\) | Structurally proved from BFS analysis |
| \([H1]\) | Weil-block-compatible embeddings | Proved on the admissible low-frequency sector |
| \([H2]\) | Strong convergence of rescaled generators | Open; primary representational target |
| \([H\text{-}w]\) | Admissibility weights \(a_q(s)\to A>0\) | Hypothesis; supported by O-series data |
| \([H\text{-}E1]\) | Uniform Poincaré/coercivity estimate | Hypothesis; accessible analytic target |
| \([C]\) | Mosco tightness / no spectral escape | Quantitatively confirmed; proof open |
Why \([H1]\) is no longer a structural obstruction
The full embedding hypothesis asks for compatibility not only with additive Hilbert structure but also with the multiplicative Weil-block decomposition. Q5a shows that this full requirement is not needed for the continuum-limit problem addressed by \(T1\)-\(T3\).
The reason is spectral: admissible test vectors are confined by the projection \(\Pi_q\) to a low-frequency sector \(|\xi|\lesssim 1\ll q\). In this regime the sinc embedding is canonical, and the distinction between full Weil-block compatibility and additive band-limited convergence becomes invisible.
Thus \([H1]\) remains an algebraic refinement in full generality, but not a logical prerequisite for the Mosco compactness problem.
Conjecture C: spectral tightness
The core analytic difficulty is the liminf side of Mosco convergence. It is equivalent to showing that energy-bounded admissible sequences cannot send spectral mass to the high-frequency boundary \(|\xi|\sim q\).
In Q5a this is formulated as: for every \(\varepsilon>0\), there exists \(R>0\) such that
\[ \sup_q \sum_{|\xi|>R} |\widehat{f}_q(\xi)|^2 < \varepsilon . \]
Numerically, the tested admissible profiles satisfy a much stronger finite-\(q\) criterion at \(R=q/3\), with spectral tail mass below \(2.1\times 10^{-5}\).
Conceptual interpretation
Q5a changes the meaning of the continuum limit. The continuum is not obtained by making a finite set of points denser. It is obtained because admissibility forms become coercive and compact enough to force precompactness of spectral profiles.
In this sense, the limiting operator \(L_\Pi\) is not imposed externally. It is read off from the stability properties of the admissible projection.
The one-dimensional operator \(L_\Pi=-A\partial_x^2\) is not claimed to be the full spacetime geometry. It is the operator-theoretic shadow of the homogeneous four-dimensional Heisenberg structure; the extraction of the effective Lorentzian geometry is the role of Q5b.
Relation to the Cosmochrony programme
Q5a sits at the point where the finite admissible fibre becomes an operator-theoretic continuum object.
The relevant chain is: Foundation establishes admissible non-injective transitions; HeisenbergStructure identifies \(F_n\simeq V_\rho\); Q5a formulates the large-\(q\) Hilbert and operator limit; Q5b extracts the four-dimensional Lorentzian geometry; Q7 and later papers analyse the bridge between admissible dimension and spatial dimension.
This makes Q5a the analytic bridge between the discrete O-series pipeline and the continuous geometric Q-series.
Current result and next step
- Established: the Hilbert limit is controlled by the sinc embedding, and \([H1]\) is proved in the admissible low-frequency sector.
- Conditionally established: representational convergence and Mosco convergence follow from the stated analytic hypotheses.
- Numerically supported: spectral tightness is strongly supported by the high-frequency tail test.
- Remaining core: prove uniform coercivity \([H\text{-}E1]\) and spectral tightness \([C]\), or replace them with a sharper structural theorem for admissible Fourier atomicity.
Reference
Jérôme Beau. Large-q Limit of the Admissible Fibre: Representational and Operator Convergence. Q5a of the Cosmochrony Spectral Geometry Programme, 2026.