Spectral Morphology Clustering and the Hilbert-Pólya Conjecture: An Exploratory Negative

Abstract

This essay investigates whether the unsupervised machine learning classification of Fast Radio Bursts by spectral morphology parameters, as reported in Sun et al.

Source Material and Tempting Analogies

The source paper analyzes 4,527 Fast Radio Bursts from the CHIME/FRB Catalog 2 using Uniform Manifold Approximation and Projection (UMAP) combined with HDBSCAN density-based clustering. The authors identify two primary clusters in the eight-dimensional parameter space (flux, fluence, pulse width, spectral index γ, spectral running r, and frequency bounds), achieving 94% recall for known repeating sources. The spectral running parameter r — describing the curvature of the log-flux versus log-frequency relation — emerges as the dominant discriminator between repeating (narrowband) and non-repeating (broadband) populations.

The Failed Spectral Bridge

The apparent resonance with the Riemann Hypothesis arises from the shared vocabulary of "spectral" analysis and the emergence of discrete binary structure from continuous parameters. One might speculate that the spectral running r resembles a spectral parameter deforming an operator, or that the two FRB populations parallel the distinction between trivial and non-trivial zeta zeros. This essay examines whether such a Hilbert-Pólya bridge can be constructed.

Assessment of the Analogy

We rate this analogy as FAILED at the level of Formal Analogy. The "spectral" parameters in the FRB paper are phenomenological descriptors of electromagnetic emission bandwidth, not eigenvalues of a self-adjoint operator. The clustering lacks the characteristic GUE spacing statistics, determinant structure, or underlying Hamiltonian dynamics required by the Hilbert-Pólya program. While the unsupervised learning reveals astrophysical structure, it provides no analytic framework connecting to the critical line.

This essay was produced by an automated research pipeline and has not been peer reviewed; conjectures herein are unproven.