UNNS Substrate Research Program · Phase Mapping Protocol · April 2026

Phase Mapping Corpus Analysis
α–μ Joint Operator Deformation

First complete execution of the Phase Mapping Protocol on real atomic and molecular physical systems. 17×17 joint operator grid across α ∈ [0.80,1.20] and μ ∈ [0.80,1.20] — 289 evaluations per system, 9,826 total. Forward composition μ(α(L)) vs reverse composition α(μ(L)) compared at every grid point.
STRUC-PERC-I v2.4.1 9,826 total evaluations 34 datasets · 17×17 grid 22 FULL · 5 TAIL · 7 HARD All commutators: zero All systems: structurally rigid
Total evaluations
9,826
34 datasets × 289 grid points
FULL datasets
22
GR=1.000 at all grid points
TAIL datasets
5
Zeeman TAIL (each dataset's GR fixed; range 0.955–0.984 across datasets)
HARD datasets
7
H, Na Zeeman, CO, HD combined
Verdict changes
0
across all 9,826 evaluations
Non-zero commutators
0
forward = reverse everywhere
Intra-grid GR variation
0.000
per dataset, across all (α,μ) points
Structurally rigid
34/34
verdict constant within each dataset across all (α,μ)

Primary Findings

Finding 1 — Universal Structural Rigidity

Every one of the 34 tested datasets satisfies Theorem 5 of the Phase Mapping Protocol (structural rigidity): for each dataset individually, the verdict V(α,μ) is constant across all tested (α,μ) points, and the giant ratio GR is identical at every grid point within that dataset. No dataset exhibits a phase transition anywhere in the tested parameter space α ∈ [0.80, 1.20] × μ ∈ [0.80, 1.20]. The verdict of a fixed ladder construction is invariant under (α,μ) deformation — though it varies across different representations of the same physical system (Finding 3). This result holds across atomic, molecular, and nuclear domains.

Finding 2 — Trivial Commutator Throughout

The structural commutator C(α,μ;L) = S(μ(α(L))) − S(α(μ(L))) is zero at every grid point for every dataset: Δ_verdict = 0, Δ_giant = 0.000000, Δ_κ = 0. All tested ladders are structurally commutative (Theorem 2 of Phase Mapping Protocol): the order in which α and μ are applied to a fixed ladder construction does not change the structural outcome. This is a consequence of structural rigidity: a ladder whose GR and verdict do not change under either operator individually cannot produce asymmetry under their composition. Conjecture 3.4 of the Structural Response manuscript (operator non-commutativity) therefore remains untested here — to observe non-commutativity the pipeline must be applied to transitional systems where single-operator response is non-trivial.

Finding 3 — Representation Is the Primary Structural Variable

The dominant source of structural variation in this corpus is not operator deformation but ladder representation. The same physical system (He, Na, HD) yields categorically different verdicts depending on how the ladder is constructed. This is the dominant signal in this dataset and provides the strongest corpus evidence yet for Theorem 7.5 of the Dual Observability manuscript (representation non-invariance). Intra-grid GR variation (within a single dataset, across all (α,μ) points): zero — GR is constant for every fixed ladder construction. Inter-representation GR variation (across different ladder constructions of the same physical system): up to 0.20 (HD combined vs HD lower: 0.800 vs 1.000). These are different comparison levels: the first is invariance under deformation, the second is sensitivity to encoding.

Finding 4 — Sub-Regime Metric Structure Within Rigid Systems

Even though all systems are verdict-rigid, their κ_conn values differ dramatically within the FULL class: CH4/CO2/H2O/NH3/O3 connect at κ < 0.3 (near-immediate), while He QM-I connects at κ = 10⁶ (maximum adaptive extension). Na QM-I connects at κ = 161,260 and Li at κ = 376,546. These differences are invisible to the verdict but constitute Theorem 3 of the Phase Mapping Protocol (hidden structural variation). The sub-regime metric structure is constant across the grid — it is a property of the ladder, not of the deformation.

Phase Space Maps — Selected Systems

Each cell is one (α,μ) grid point. Colour: green=FULL, amber=TAIL, red=HARD. All maps are monochromatic — no system transitions across the grid.

Phase maps: V(α,μ) for selected systems — 17×17 grid each He (QM-I) FULL · all 289 points α → μ → 0.8 1.2 0.8 1.2 He (Zeeman) TAIL · all 289 points α → μ → 0.8 1.2 0.8 1.2 H (all repr.) HARD · all 289 points α → μ → 0.8 1.2 0.8 1.2 Na (QM-I) FULL · all 289 points α → μ → 0.8 1.2 0.8 1.2 Na (Zeeman) HARD · all 289 points α → μ → 0.8 1.2 0.8 1.2 HD combined HARD · all 289 points α → μ → 0.8 1.2 0.8 1.2 All maps monochromatic: for each fixed ladder construction, verdict is constant under (α,μ) deformation FULL TAIL HARD No phase transition detected in any system

Commutator Analysis — C(α,μ;L)

The structural commutator measures whether the order of operator application matters: C(α,μ;L) = S(μ(α(L))) − S(α(μ(L))). A non-zero commutator signals operator-order sensitivity.

Result: Globally Trivial Commutator

Across all 9,826 (forward, reverse) evaluation pairs: Δ_verdict = 0, Δ_giant = 0.000000, Δ_κ = 0 in every single case. The structural commutator is identically zero throughout the corpus. This means: for all tested ladders, the joint (α,μ) deformation is structurally commutative — the order in which α and μ are applied does not change the structural outcome.

This is consistent with (but stronger than) the marginal-sweep operator separation result: if GR does not change under either α alone or μ alone, the composition cannot produce an asymmetry either. The trivial commutator is thus a consequence of structural rigidity, not an independent finding. The combination is degenerate in the sense of Corollary 1 of the Phase Mapping Protocol.

L α(L) α(L) μ(α(L)) μ(α(L)) μ(L) μ(L) α(μ(L)) α(μ(L)) = = Same verdict Same GR ΔC = 0 everywhere

Representation Sensitivity — The Dominant Structural Variable

For any fixed ladder construction, the structural class is invariant under (α,μ) deformation (Finding 1). But different ladder constructions of the same physical system can yield categorically different classes. The dominant source of verdict variation across the corpus is representation — not constant deformation — and this corpus substantially extends the helium representation-split first identified in the single-constant sweep.

Helium

4 representations — 2 distinct classes

QM-I preprocessed (n=1999)FULL
Gap structure QM-I (n=1999)FULL
Spectrum QM-I (n=1684)FULL
Zeeman ladder (n=1999)TAIL GR=0.958
Singlet Zeeman (n=1999)TAIL GR=0.984
Triplet Zeeman (n=1999)TAIL GR=0.966

QM-I ladders uniformly FULL; Zeeman-split ladders uniformly TAIL. The fine structure exposed in Zeeman splitting reveals permanent tail isolation.

Hydrogen

4 representations — all HARD

QM-I preprocessed (n=160)HARD GR=0.850
Gap structure QM-I (n=129)HARD GR=0.814
Spectrum QM-I (n=211)HARD GR=0.934
Zeeman ladder (n=1999)HARD GR=0.975

Representation-invariant HARD: all four constructions return HARD at every (α,μ) point. H is robustly HARD regardless of encoding. Principal quantum number shell structure dominates every representation.

Sodium

4 representations — 2 distinct classes

QM-I preprocessed (n=1385)FULL κ=161k
Gap structure QM-I (n=1038)FULL κ=161k
Spectrum QM-I (n=836)FULL κ=207k
Zeeman ladder (n=1999)HARD GR=0.922

Most extreme split: QM-I → FULL at large κ, Zeeman → HARD. Zeeman sub-structure introduces magnetic-field-split outlier gaps that fragment the vulnerability graph severely.

HD (Hydrogen Deuteride)

3 representations — 2 distinct classes

Lower levels (n≈290)FULL κ=0.7
O-branch lower (n=244)FULL κ=0.6
Combined levels (n=1999)HARD GR=0.800

Merging sub-ladders creates permanent fragmentation. The combined ladder incorporates transitions spanning very different energy scales; the resulting extreme gap ratios prevent backbone formation.

CO (Carbon Monoxide)

Conflicting representations across corpora

This corpus (molecular_co)HARD GR=0.937
Prior STRUC-PERC corpusFULL κ=714,804

Cross-corpus discrepancy: CO returns HARD here but FULL in the previous μ-deformation corpus. Different ladder constructions (rovibrational level selection, gap preprocessing) produce opposite verdicts. Requires canonical ladder investigation.

Complete System Table — All 34 Datasets

DatasetDomainVerdict (all γ)GR at (1,1)κ_connn (gaps)Phase stableC(α,μ)
calcium_zeeman_ladder Atomic (Zeeman) FULL 1.0000 296,875 1999 0
iron_zeeman_ladder Atomic (Zeeman) FULL 1.0000 5,960 1999 0
helium_QM1_preprocessed Atomic (QM-I) FULL 1.0000 1,000,000 1999 0
helium_gap_structure_QM1 Atomic (QM-I) FULL 1.0000 1,000,000 1999 0
helium_spectrum_QM1 Atomic (QM-I) FULL 1.0000 271,999 1684 0
heliumII_QM1_preprocessed Atomic (QM-I) FULL 1.0000 1,000,000 589 0
heliumII_gap_structure_QM1 Atomic (QM-I) FULL 1.0000 1,000,000 441 0
heliumII_spectrum_QM1 Atomic (QM-I) FULL 1.0000 329,031 296 0
lithium_QM1_preprocessed Atomic (QM-I) FULL 1.0000 376,546 460 0
lithium_gap_structure_QM1 Atomic (QM-I) FULL 1.0000 376,546 366 0
lithium_spectrum_QM1 Atomic (QM-I) FULL 1.0000 420,033 277 0
sodium_QM1_preprocessed Atomic (QM-I) FULL 1.0000 161,260 1385 0
sodium_gap_structure_QM1 Atomic (QM-I) FULL 1.0000 161,260 1038 0
sodium_spectrum_QM1 Atomic (QM-I) FULL 1.0000 206,721 836 0
gold_zeeman_ladder Atomic (Zeeman) TAIL 0.9705 none 1999 0
helium_zeeman_ladder Atomic (Zeeman) TAIL 0.9575 none 1999 0
helium_singlet_zeeman_ladder Atomic (Zeeman) TAIL 0.9840 none 1999 0
helium_triplet_zeeman_ladder Atomic (Zeeman) TAIL 0.9655 none 1999 0
silver_zeeman_ladder Atomic (Zeeman) TAIL 0.9555 none 1999 0
hydrogen_zeeman_ladder Atomic (Zeeman) HARD 0.9750 none 1999 0
hydrogen_QM1_preprocessed Atomic (QM-I) HARD 0.8500 none 160 0
hydrogen_gap_structure_QM1 Atomic (QM-I) HARD 0.8140 none 129 0
hydrogen_spectrum_QM1 Atomic (QM-I) HARD 0.9336 none 211 0
sodium_zeeman_ladder Atomic (Zeeman) HARD 0.9220 none 1999 0
molecular_CH4_ladder Molecular FULL 1.0000 0.2 1999 0
molecular_CO2_ladder Molecular FULL 1.0000 0.1 1999 0
molecular_H2O_ladder Molecular FULL 1.0000 0.3 1999 0
molecular_NH3_ladder Molecular FULL 1.0000 0.1 1999 0
molecular_O3_ladder Molecular FULL 1.0000 0.3 1999 0
HD_lower_levels Molecular (HD) FULL 1.0000 0.7 385 0
HD_O_lower_levels Molecular (HD) FULL 1.0000 0.6 244 0
nuclear_level_ladders Nuclear FULL 1.0000 2,578 1999 0
molecular_co_ladder Molecular HARD 0.9370 none 1999 0
HD_combined_levels Molecular (HD) HARD 0.8004 none 1999 0

κ_conn Distribution — Hidden Metric Structure Within FULL Class

All FULL systems share GR=1.000 and a constant verdict across the entire α×μ grid. Yet their connectivity thresholds span five orders of magnitude — the same sub-regime metric structure established in the prior corpus. This is deformation-invariant: κ_conn does not change when α or μ varies.

κ_conn at physical value — FULL systems (log scale) · invariant across (α,μ) grid CO2 0.1 NH3 0.1 CH4 0.2 H₂O 0.3 O3 0.3 HD O-lower 0.6 HD lower 0.7 Nuclear 2577 Fe Zeeman 5960 Na QM-I 161260 Na spectrum 206721 He spectrum 271999 Ca Zeeman 296875 HeII spectrum 329031 Li QM-I 376546 Li spectrum 420033 He QM-I 1000000 HeII QM-I 1000000 0.1 1 10 100 10³ 10⁴ 10⁵ 10⁶ κ_conn → Molecular (immediate) Nuclear/Fe (moderate) Atomic QM-I (high) Atomic extreme

Framework Implications

Theorems Confirmed (Phase Mapping Protocol)

TheoremStatementStatus
Theorem 5Structural Rigidity: V constant over Ω, C=0CONFIRMED — all 34 datasets
Corollary 1Degenerate Phase Space: Φ(α,μ) = V₀ everywhereCONFIRMED — all 34 datasets
Theorem 2Structural Commutativity: C(α,μ;L) = 0CONFIRMED — all 34 datasets
Theorem 3Hidden Structural Variation: Δ_V=0 but internal metric differsCONFIRMED — κ_conn varies 10⁵× within FULL class
Theorem 4Phase Transition: V changes discontinuouslyNOT OBSERVED — no transitions in any tested system
Corollary 2Non-Commutative Region: C(α,μ;L) ≠ 0NOT OBSERVED — commutator identically zero

Relation to Structural Response Manuscript

Operator Non-Commutativity Conjecture (Conjecture 3.4)

The joint (α,μ) sweep provides the first direct test of Conjecture 3.4. Result: trivial commutator throughout. This is consistent with the conjecture being false for structurally rigid systems — if the verdict and GR are deformation-invariant under each operator separately, they will remain so under any composition. The conjecture may only be testable on systems with non-trivial single- operator response (i.e. transitional systems in the ²⁸Si / ²³⁸U class). Those systems are not in this corpus.

Representation Non-Invariance (Theorem 7.5)

This corpus provides the strongest corpus evidence yet for representation non-invariance: Na QM-I is FULL while Na Zeeman is HARD; He QM-I is FULL while He Zeeman is TAIL; HD lower is FULL while HD combined is HARD; CO here is HARD while CO in the prior μ-corpus is FULL. The structural class is not a property of the physical system alone — it is a joint property of the system and its ladder encoding. The canonical ladder problem (Proposition 7.3) is the critical open problem for this corpus.

Next Steps

To test Conjecture 3.4 (operator non-commutativity), the pipeline should be applied to transitional systems — specifically ²⁸Si and ²³⁸U nuclear isotopes, which showed class-change behaviour in the single-operator α sweep. These systems occupy the transitional class Ctrans(α) and may exhibit non-trivial commutator behaviour under joint (α,μ) deformation. Geoid harmonics at physical α are also priority targets.

Instrument: STRUC-PERC-I v2.4.1 · Stage 12.6 theory-consistency patch · Protocol: Phase Mapping Protocol v1.0 · forward: μ(α(L)), reverse: α(μ(L)) · Grid: α ∈ [0.80,1.20], μ ∈ [0.80,1.20], 17×17=289 points each · Corpus: 34 datasets · 9,826 total evaluations · atomic + molecular + nuclear · Datasets: H (4 repr.), He (6 repr.), HeII (3 repr.), Li (3 repr.), Na (4 repr.), Ca/Fe/Au/Ag Zeeman, CH4/CO2/H2O/NH3/O3/CO molecular, HD (3 repr.), nuclear levels · Date: April 2026 · UNNS Substrate Research Program

New Datasets — Silicon (Condensed Matter) & Uranium-238

Three silicon representations from the Materials Project condensed-matter database and two uranium-238 ladder constructions (full ENSDF level scheme + top-levels only) add 3 new phase-mapped systems and 2 static STRUC-PERC-I single runs to the corpus.

Silicon — Materials Project (3 representations, 867 grid evaluations)

All three Si constructions: FULL — structurally rigid across entire α×μ grid

The density ladder (n=41 gaps), formation energy ladder (n=42 gaps), and table export (n=271 gaps) all return FULL_PERCOLATION at every one of their 289 grid points. GR = 1.000 throughout. Commutator identically zero. Rigid class: confirmed.

Si Representationn gapsVerdict (all 289 pts)κ_connPhase stableC(α,μ)
Density ladder (Materials Project) 41 FULL 2.000 0
Formation energy ladder 42 FULL 81.4 0
Table export (full phase table) 271 FULL 56,394 0

Cross-domain representation split — Si condensed matter vs Si-28 nuclear

Silicon in the Materials Project (condensed matter, phase/formation-energy ladder) is FULL and rigid across the entire α×μ grid. Silicon-28 in the prior nuclear constants deformation corpus was transitional HARD — HARD at γ_α ≤ 0.90, FULL at the physical value. Same chemical element, categorically different structural behaviour depending on whether the ladder is built from condensed-matter phase data or nuclear γ-level spectroscopy. This is the clearest cross-domain representation split in the corpus: the same element (Si) occupies FULL (condensed matter) vs transitional HARD (nuclear) depending on the physical context of the ladder construction.

Sub-regime κ_conn variation within Si FULL class

The three FULL Si representations span κ_conn = 2 (density, immediate) to 56,394 (table export, high delay) — a factor of 28,000 within the same element under the same domain classification. As with the prior corpus, the FULL verdict is informationally incomplete: the metric depth (κ_conn) varies by orders of magnitude within FULL even for the same element. All three κ_conn values are constant across the entire 289-point α×μ grid.

Si phase maps — 3 representations · 17×17 grid each · all FULL Si density (κ=2) FULL · 289/289 Si formation E (κ=81) FULL · 289/289 Si table export (κ=56k) FULL · 289/289 All Si constructions: monochromatic FULL — verdict and GR constant at every (α,μ)

Uranium-238 — Two Ladder Constructions (Static STRUC-PERC-I Runs)

Two U-238 ladder constructions are evaluated here as single-point static runs (not phase-mapped across the full α×μ grid). They extend the representation analysis with a new split within the same isotope.

ENSDF full level scheme (1,112 gaps) — TAIL · GR = 0.9829

The full ENSDF U-238 nuclear level scheme (1,113 elements, 1,112 gaps) returns TAIL_FRAGMENTATION with GR = 0.9829 and 14 permanently isolated vertices. 34 outlier gaps (3.1% of all gaps) with max/median ratio = 2.48×10⁹ dominate the tail. These are single γ-transitions far above the bulk level spacing that cannot merge with the main backbone within the tested κ range. Adaptive extension pushes GR to 0.9991 (2 components remaining) — extremely close to FULL but not achieved. Theorem 1 of the revised PRP does not apply (TAIL class). Fully admissible.

All levels (top 5 only, 4 gaps) — FULL · κ_conn = 0.56

A minimal U-238 construction using only the top-level transitions (5 elements, 4 gaps) returns FULL_PERCOLATION with κ_conn = 0.56 and zero outliers. With only 4 gaps in a relatively uniform energy range, the gap structure connects almost immediately.

²³⁸U ENSDF — giant ratio GR(κ) across scale grid + adaptive extension 0 0.5 0.75 0.90 1.00 Giant Ratio (GR) κ=1 ← standard grid adaptive extension → GR=0.983 0.01 0.1 1 10 10³ 10⁶ κ (scale) → Standard grid (κ ≤ 1) Adaptive extension

²³⁸U representation split — TAIL (ENSDF) vs FULL (top-levels) vs FULL at physical in constants corpus

Three U-238 ladder constructions produce three different structural outcomes: (1) Full ENSDF level scheme (1,112 gaps) → TAIL, GR = 0.9829, 34 outlier γ-transitions with max/median = 2.48×10⁹. (2) Top 5 levels only (4 gaps) → FULL, κ_conn = 0.56 — the coarse top-level structure connects immediately with no outliers. (3) Prior constants deformation corpus → FULL at physical α, transitional HARD at γ_α = 1.02. The three constructions map to three distinct structural regimes. U-238 joins He, Na, HD, CO, and Si as an element demonstrating categorical representation-sensitivity.

Updated Corpus Totals

Total datasets (phase-mapped)
40
+3 Si · note: B3 Zeeman counted separately below
Total grid evaluations
11,560
+867 Si evaluations
FULL at physical value
28
+3 Si (all FULL)
TAIL at physical value
5
B1 Zeeman n=1999 constructions
HARD at physical value
7
unchanged
Intra-dataset rigidity
40/40
every dataset: verdict constant within its (α,μ) grid
Non-zero commutators
0
across all 11,560 evaluations
Representation-split elements
7
He Na H HD CO Si U-238

Key new finding — Cross-domain element split (Si)

Si condensed matter (Materials Project density/formation-energy/table) → FULL, rigid. Si-28 nuclear (ENSDF, constants deformation corpus) → transitional HARD at low α. The element silicon occupies different realizability classes in different physical domains. The structural class is a joint property of the element, its domain, and its ladder encoding — not of the element alone.

New datasets added: Si_density_ladder · Si_formation_energy_ladder · Si_table_export (Materials Project, phase-mapped 17×17) · ensdf_238u (ENSDF full level scheme, static run) · 238u_all_levels (top 5 levels, static run) · Date: 2026-04-13 · UNNS Substrate Research Program

FULL_OUTPUT Batch — 7 New Domains, 55 New Datasets

A major corpus extension covering six new physical domains and an updated Zeeman atomic dataset. 55 new datasets across Atmosphere (ERA5), CMB (Planck R3.01), Cosmic Web (μ-sweep), Crystallography (Materials Project), Geoid (Earth/Mars/Moon), Solar (GOES-XRS), and U-238 sub-constructions — plus 9 updated Zeeman ladders.

New datasets
55
7 new domains
New grid evaluations
14,680
+14,680; verified total: 22,817
New FULL datasets
46
incl. new Zeeman, Solar, Atm.
New HARD datasets (cryst.)
9
per_atom / per_formula_unit
Non-zero commutators
0
all new datasets: trivial C
Structurally rigid
55/55
Theorem 5: universal

1 · Zeeman Atoms — Complete Reversal of Previous TAIL/HARD Verdicts

7 of 9 Zeeman atoms changed class — largest cross-batch representation shift yet

The previous Zeeman corpus (n=1999 gaps each) returned TAIL for Au, He, singlet He, triplet He, Ag and HARD for H, Na. The new Zeeman corpus (n=39–124 actual level gaps) returns FULL for all 9 atoms. Zero TAIL, zero HARD. Same physical systems, completely different structural class — determined entirely by ladder construction. This is the largest single-batch representation reversal in the corpus.

AtomOld verdict (n=1999)Old GRNew verdict (actual n)New GRNew κ_connNew n
CalciumFULL1.000FULL1.00032242
GoldTAIL0.971FULL1.000652
HeliumTAIL0.958FULL1.00011,18940
Helium singletTAIL0.984FULL1.00013,70939
Helium tripletTAIL0.966FULL1.00011,20540
HydrogenHARD0.975FULL1.000187,01439
IronFULL1.000FULL1.00010124
SilverTAIL0.956FULL1.000839
SodiumHARD0.922FULL1.00032939

Mechanism: ladder size determines class

The previous corpus inflated all ladders to n=1999 gaps — which for sparse Zeeman level sets introduced large structural artifacts (zero-gaps, repeated values, or bulk padding that created extreme outlier-to-bulk ratios). The new corpus uses the actual level count (39–124 gaps). With a compact realistic ladder, the Zeeman-split levels connect well within the tested κ range. Hydrogen now returns FULL at κ_conn = 187,014 — the second-highest κ_conn in the entire corpus, just below He QM-I at 10⁶. The large connectivity delay reflects the genuine principal-quantum-number shell structure, but with the actual level count the backbone forms. Na Zeeman similarly recovers to FULL at κ=329. The canonical ladder problem is the central open question this result underscores.

2 · Crystallography — Cell Volume Always FULL, Normalised Volumes Often HARD

Clean structural pattern: the normalisation axis determines the class

cell_volume (total unit cell volume per phase) → FULL for all 6 compounds tested (BaTiO3, Fe, KNbO3, SiO2, TiO2, ZrO2). volume_per_atom → HARD for 5 of 6 (SiO2 exception). volume_per_formula_unit → HARD for 4 of 6 (SiO2 and BaTiO3 exceptions). The cell-volume ladder encodes one structural value per phase; the per-atom and per-formula-unit normalizations produce specific volume ratios that, for most perovskite/oxide compounds, generate isolated extreme outliers preventing backbone formation. SiO2 is the exception: all three representations return FULL, suggesting its phase volume structure is more uniform across normalizations.

Compoundcell_volumeκ_connvolume_per_atomvolume_per_formula_unitStructure type
SiO2 FULL1.0 FULL κ=10.0 FULL κ=5.6 silica (quartz/cristobalite)
BaTiO3 FULL0.7 HARD GR=0.800 FULL κ=0.7 perovskite
Fe FULL1.0 HARD GR=0.750 HARD GR=0.750 BCC iron
KNbO3 FULL1.0 HARD GR=0.900 HARD GR=0.900 perovskite
TiO2 FULL0.7 HARD GR=0.714 HARD GR=0.714 rutile/anatase
ZrO2 FULL1.0 HARD GR=0.857 HARD GR=0.857 zirconia

All 18 crystallography datasets: structurally rigid, trivial commutator

Despite the FULL/HARD split across representations, every dataset returns the same verdict at every (α,μ) grid point. The crystallographic phase-volume ladder structure is completely insensitive to constant deformation. HARD here is representation-intrinsic, not deformation-induced.

3 · New Domains — Atmosphere, CMB, Geoid, Cosmic Web, Solar

Atmosphere (ERA5) · 8 datasets

All FULL · κ_conn = 0.7–2.0 · immediate connectivity

ERA5 250hPa zonal wind: global longitude sectors, latitude bands, jet-region aggregates — 8 representations of upper-tropospheric jet-stream structure. All FULL, rigid, trivial commutator. κ_conn ≤ 2 for all constructions — among the lowest in the entire corpus. Atmospheric wind structure has near-immediate gap-graph connectivity under any (α,μ) deformation.

latband absmaxFULL κ=0.7
latband signedmeanFULL κ=1.0
global lonsector / top-12 jetFULL κ=2.0
CMB Planck R3.01 · 3 datasets

All FULL · κ_conn = 1.0–7.9

Planck 2018 R3.01 power spectra (TT, TE, EE) with the full multi-frequency Planck dataset. Note the dramatic κ_conn change from the prior CMB results (TT=134k, EE=859k) — these R3.01 ladders are much smaller (n=54–499 vs n=thousands) producing lower κ_conn. Same physical signal, different ladder construction, same FULL verdict but very different connectivity depth.

CMB EE (n=54)FULL κ=1.0
CMB TE (n=499)FULL κ=1.0
CMB TT (n=399)FULL κ=7.9
Geoid Earth · Mars · Moon

All FULL · μ-only sweeps (κ_conn = 0.7–6.3)

First μ-sweep of geoid harmonic ladders. EIGEN-6C4 Earth (κ=1.0, n=181), JGM85 Mars (κ=0.7, n=499), AIUB-GRL350A Moon (κ=6.3, n=142) — all FULL across 21 μ grid points at fixed physical α. Previously only tested at α=1.10 and 1.20 (non-physical, HARD). At physical α the geoids are FULL and μ-stable. The prior HARD result is confined to amplified α beyond the physical range, consistent with the Realizability Anchoring Conjecture.

Cosmic Web · 7 datasets · μ-sweep

All FULL · μ-stable · κ_conn = 28–15,658

DESI, SDSS, 2MRS cosmic web orientation ladders — μ-only sweep (α=1.0 fixed, μ ∈ [0.95,1.05]). All 7 constructions FULL and μ-stable. κ_conn spans 27.9 (SDSS) to 15,658 (DESI rotated sample) — reflecting different structural density of the cosmic web representations. The DESI full synthetic (n=666, κ=5,908) and DESI rotated (n=1999, κ=15,658) show large connectivity delays consistent with the sparse void-dominated cosmic web structure.

Solar — GOES-XRS · First Solar Domain

FULL · κ_conn = 16.6 · n = 7 gaps · 289 grid points

First solar domain in the corpus. GOES X-ray solar flux ladder (7 gaps from 8 X-ray energy level values) — FULL across the entire 17×17 (α,μ) grid. κ_conn = 16.6. The solar X-ray spectral gap structure connects at moderate κ, rigid across all constant deformations. This extends the zero-violation, FULL-at-physical result to the solar electromagnetic domain for the first time.

4 · Updated Corpus Totals

Total datasets (all batches)
93
B1(34) + B2(3) + B3(56)
Total grid evaluations
22,817
verified count across all batches
FULL at physical value
72
includes B3 Zeeman reversal
HARD at physical value
16
+9 crystallography (B3)
TAIL at physical value
5
B1 Zeeman n=1999 constructions only
Non-zero commutators
0
across all 22,817 evaluations
Intra-dataset rigidity
93/93
verdict constant within each dataset
Physical domains covered
11
+Solar, +Crystallography

Three universal results — now confirmed across 22,817 evaluations, 11 physical domains

1. Structural rigidity (Theorem 5): For each fixed ladder construction, the verdict V(α,μ) and giant ratio GR(α,μ) are constant across the entire tested parameter space — zero intra-dataset variation at any (α,μ) point. No exceptions across 93 datasets, 22,817 grid evaluations, 11 physical domains. No phase transition has been observed in any tested system under joint (α,μ) deformation. This invariance holds within each fixed construction; different constructions of the same physical system can yield different verdicts (see point 3).

2. Trivial commutator: The structural commutator C(α,μ;L) = 0 for every tested ladder, at every grid point. For any fixed ladder, the order in which α and μ are applied never changes the structural outcome. Operator non-commutativity has not been observed in any tested system.

3. Representation is the primary structural variable: While constant deformation never changes the verdict of a fixed ladder construction, the choice of ladder construction itself does — sometimes categorically. The dominant source of verdict variation across the corpus is representation, not deformation. The Zeeman reversal (7 class changes between batches), the crystallography normalization split (cell_volume vs per_atom), and the cross-domain Si and U-238 splits all confirm that realizability class is a joint property of physical system + ladder encoding, not of the physical system alone. The canonical ladder problem remains the highest-priority open question.

New datasets: Atmosphere ERA5 (8) · Atomic Zeeman updated (9) · CMB Planck R3.01 (3) · Cosmic Web DESI/SDSS/2MRS μ-sweep (7) · Crystallography Materials Project (15 valid + 3 Al2O3 errors) · Geoid Earth/Mars/Moon μ-sweep (3) · Solar GOES-XRS (1) · U-238 sub-constructions (5, deduplicated) · Al2O3: 3 datasets produced engine errors (no result) — excluded · Date: 2026-04-14 · UNNS Substrate Research Program