UNNS Substrate Research Program · unns.tech · April 2026
UNNS Substrate Research Dashboard
Local Geometry of Realizability Boundaries
Core manuscript · validation corpus · cross-system synthesis · STRUC-PERC-I engine
[ Atomic Phase Landscape ] │ [ Helium Validation ][ CORE MANUSCRIPT ][ 28Si Validation ] [ STRUC-PERC-I ENGINE ] │ [ Raw / Output Data ]
WORKING MANUSCRIPT
16 DIRECT CORPUS RUNS
6 ELEMENTS · 5 REGIMES
STRUC-PERC-I v2.4.0
UNNS SUBSTRATE RESEARCH PROGRAM · WORKING MANUSCRIPT · APRIL 2026
Local Geometry of Realizability
Boundaries in the UNNS Substrate
FROM CONNECTIVITY MARGIN TO BOUNDARY DISTANCE
Defines realizability charts and decisive coordinate systems, proves that realizability-class boundaries are locally codimension-1 C¹ hypersurfaces, establishes the bi-Lipschitz equivalence of the connectivity margin m(L) with the structural boundary distance d_∂C(L), and derives the Local Maximum-Margin Canonicalization Theorem. Includes §2.6 (ladder construction protocols), §8.7 (cross-system atomic phase landscape), and Appendix D (protocol specification with exact CSV column mappings).
THEOREM 4.3
Local Boundary Hypersurface
Near every regular boundary point, ∂C is a codimension-1 C¹ hypersurface in a decisive chart.
THEOREM 6.4
Local Margin Monotonicity
d_∂C(L₁) < d_∂C(L₂) ⟹ m(L₁) < m(L₂) near any regular boundary point.
COROLLARY 6.3
Order Equivalence
m(L) is locally order-equivalent to d_∂C(L); the ordering is chart-independent.
THEOREM 7.2
Local Canonicalization
All margin-maximizing encodings in a regular local regime belong to a single realizability class.
28
PAGES
11
THEOREMS & LEMMAS
5
APPENDICES
4
PROTOCOL FAMILIES
0
USL VIOLATIONS
The core manuscript: proves local boundary hypersurface structure, bi-Lipschitz margin-distance equivalence, local monotonicity theorem, and maximum-margin canonicalization. 28 pages with §2.6 (protocol definitions), §8.7 (cross-system synthesis), and Appendix D (exact CSV column specifications for all protocol families).
STRUCTURAL PERCOLATION ENGINE
STRUC-PERC-I v2.4.0 · PRP PERCOLATION ANALYZER
Full Percolative Realizability Principle (PRP) analyzer implementing the pairwise vulnerability graph G_κ(L) with IQR-scaled ε, τ-floor, and adaptive extension. Computes realizability class, connectivity margin m(L), decisive coordinates (tailDom, giant ratio, κ_conn), and structural metrics for any admissible ladder.
All empirical results in this dashboard are produced using this instrument.
81
CORPUS RUNS
16
ATOMIC RUNS
0
USL VIOLATIONS
Browser-based instrument — no installation required. Upload any CSV ladder, run the PRP percolation analysis, and download output ZIP (summary.json + layers.csv + components.csv + outliers.json). All corpus runs in this dashboard were produced using this exact instrument.
⬇ Input Data
atoms.zip ↓ Download
Raw atomic CSV sources: 24 files covering H, He, Li, Na, Fe, Ag, Au in QM-I (spectrum · preprocessed · gap-structure) and Zeeman variants. Column schemas verified and documented in Appendix D of core manuscript.
⬆ Instrument Outputs
atoms_output.zip ↓ Download
STRUC-PERC-I v2.4.0 output packages: summary.json + layers.csv + components.csv + outliers.json per run. Validated against raw CSVs: all n_elements, class verdicts, giant ratios, and tail-dominance values confirmed.
REPRODUCIBILITY STATEMENT
All results are reproducible under the protocol definitions in Appendix D of the core manuscript. The generic adapter pipeline (parse all numeric cells → sort → deduplicate → drop non-finite) produces internally consistent outputs aligned with the raw CSV sources. Canonical domain protocols (QM-I: energy_cm-1 column; Zeeman: LevelB_cm1 column) are specified for publication-grade physics interpretation.
UNNS SUBSTRATE RESEARCH PROGRAM · unns.tech · LOCAL GEOMETRY BUNDLE · APRIL 2026  ·  Version 1.0