Differentiable Lie Sector Selection in the Symmetrized Sedenion Khanhnion Model: Spectral Derivatives, Pathwise Sensitivity, and Optimal-Loss Curvature
This paper studies differentiable selection of the least-squares maximal rotation-invariant Lie sector in the standard-real two-channel symmetrized sedenion Khanhnion model. Along differentiable covariance paths with simple top eigenvalues, it derives the selected line-projector and sector-projector derivatives using classical reduced-resolvent formulas. Exact projector-speed identities yield integrated pathwise sensitivity bounds. For the optimal squared projector loss, the paper gives its first variation, proves concavity on covariance pairs, and computes its nonpositive second variation along affine paths. An explicit two-dimensional covariance block demonstrates inverse-eigengap sensitivity and diverging curvature as the top gap closes. The spectral derivative framework is explicitly attributed to established perturbation theory. The PDF includes the author-provided Khoa Toán Khánh emblem on the first page, complete proofs, Harvard-style references, an AI-assistance disclosure, and an explanatory diagram on the final page. No priority, bracket-preserving transport, physical dynamics, or computational advantage is claimed. Keywords:Khanhnion; differentiable sector selection; eigenprojector derivative; reduced resolvent; eigengap; pathwise sensitivity; optimal-loss curvature; symmetrized sedenions.
Authors
- Kim Khanh Nguyen (ORCID: https://orcid.org/0009-0006-9191-1441)
Institutions
- University of Khanh Hoa (VN)
Publication Details
- Journal
- Zenodo (CERN European Organization for Nuclear Research)
- Published
- 2026-11-19
- DOI
- https://doi.org/10.5281/zenodo.24359167
- Primary Topic
- Matrix Theory and Algorithms
- Type
- preprint