Universe OS V10-B (Exponent-Only) Proof Core Preprint (v1.2)

Universe OS V10-B (Exponent-Only) — Proof Core Preprint (v1.2) This record archives a “Proof Core” snapshot of the Universe OS V10-B (Exponent-Only) model. The purpose is to preserve a reproducible, definition-frozen demonstration that particle masses can be produced with (almost) no free parameters and without continuous fitting. Core policy (non-negotiable): Exponent-only mass rule. No linearization of the exponent is allowed. The model uses a fixed constant set and discrete structure rules. Any extension that introduces additional tunable degrees of freedom is excluded from the Proof Core of this release. Mass rule (exponent-only):m = (M_base / N_proj^k) * S Fixed constants (V10-B) are defined by: N_proj = pi^2 + e^2 + (pi/4) Phi_B = e^2 + (pi/15) alpha_geom^{-1} = N_proj * Phi_B Delta_Res = alpha_geom^{-1} - alpha_exp^{-1} M_base = 512 * Delta_Res (GeV) This preprint documents the minimal stable branch set (Lepton / Pseudoscalar / Vector / Nucleon) and provides checksum targets (e, mu, pi±, n, p, Delta m_np) for restoration verification under the same constant set. Excitations are handled without continuous tuning via dyadic quantization (W-Flag): Delta k = n/128 (n is a non-negative integer) m_exc = m_ground * N_proj^(Delta k) For high-excitation regions where experimental labels may overlap or mix, the record documents an audit discipline (“anchors vs soft references”) and a diagnostic 1-bit flag (X-Flag). X-Flag is meta-gated only: it may be used only when external meta-information indicates overlap/mixing, and it must not be toggled to reduce residuals. Scope note:This is an archival snapshot for DOI registration and reproducibility of the Proof Core. It is not a finalized peer-reviewed theory paper. Files included in this record: [RENAME THIS] UOS-V10B-ProofCore-Preprint-v1.2.pdf Preprint (PDF) compiled from the v1.2 LaTeX source. README_ZENODO_v1_2.md File list and usage notes for this record. uos_v10b_zenodo_dataset_v1_2.zip Reproducibility dataset bundle (reference Python implementation, registries/specs, and example outputs used for auditing/restoration).