Translation-free exceptional topology

Date: 26 October 2026 Status: active research arm; real-space topology is established mathematics, exceptional enrichment is a gated construction programme Seed result: rank-24 bulk collision and boundary-enrichment gate Evidence: arm ledger Certificate: rank-24 lattice verifier

Translation-free seed: disordered real-space QWZ Bott verifier

Structural cross-check: jittered-point Bott/localizer verifier

First exceptional layer: translation-free E8 edge protocol

Derived exceptional layer: translation-free E8 coupled-wire and held-out tomography protocol

Few-body surrogate: E8 bounded-degree clock hardware and gap certificate

Continuum promotion: fixed-physics clock-to-rotor convergence

Constitutive join: E8 metric calibration gate

Nonlinear solver: E8 adaptive kinetic-shell compilation

Nonlinear reconnaissance: 117,361-state joined-rotor cutoff audit

Restarted solve: 340,321-state Davidson promotion audit

Truncation firewall: E8 operator-completeness gate

Plateau candidate: Cartan-kernel shell-10 spectroscopy gate

Promoted local result: Cartan-kernel compact-cell gap

Parameter stability: Cartan-kernel coupling-ratio spine

Nonscalar stability: unequal-channel amplitude perturbation

Finite perturbation stencil: two non-collinear amplitude directions

Metric stability: positive kinetic-metric perturbation

Cross-perturbation closure: mixed amplitude-plus-metric corner

First intercell result: projected two-cell coupling

Projection audit: two-cell local-root convergence

Finite width stability: projected-chain width/root grid

Boundary comparison: open versus periodic projected chains

Reusable local interface: eight-root operator dictionary

Material bridge: quasicrystal backplane experiment

Rank-24 inference: current and defect detector ladder

Research question

What remains topological when translation symmetry, the Brillouin zone and a preferred crystalline presentation disappear, and how much additional boundary information is required before an \(E_8\), Leech or Monster description becomes physically identifying?

The arm treats an amorphous material as a witness that motivates this question, not as the sought endpoint. Its carrier is a local operator system

where \(X\) is a point set or locality graph, \(H\) is a finite-range or rapidly decaying Hamiltonian, and \(P\) is a spectral or mobility projector. The translation-free invariants are real-space indices: Bott classes, local Chern or winding markers, spectral-localizer signatures, Fredholm indices and their bulk--boundary pairings.

The programme ladder

local point set / graph + Hamiltonian
        ↓ locality and mobility/spectral gap
real-space projector or unitary K-class
        ↓ bulk-boundary index
invertible or intrinsically ordered 2+1d bulk
        ↓ chiral boundary data
E8 edge unit, c-=8
        ↓ stack/glue three units
rank-24 c=24 boundary family
        ↓ distinguish root/current content
E8^3 / Niemeier / Leech alternatives
        ↓ inversion orbifold plus twisted sector
Moonshine module V^natural
        ↓ defect, junction and OPE reconstruction
Monster-enriched boundary memory

Every arrow is a gate. The diagram is not a claim that disorder generates the Monster or that every amorphous Hall response is a topological phase.

First exact result: topology is coarser than the boundary algebra

The \(E_8\) Cartan form is positive, even and unimodular. Three copies and the Leech Gram form therefore share the coarse data

That detector cannot distinguish them. Their fine lattice data do:

Thus the same invertible bulk response can admit sharply different chiral boundary algebras. The Leech inversion orbifold adds still finer twisted and defect data and yields \(V^\natural\). This is the new arm's organizing theorem:

Topology supplies a stable coarse return class; exceptional lattices, orbifolds and Monster defects are successive enrichments of its boundary memory.

The existing corpus no-go remains essential: a pure topological carrier can forget the ultraviolet projector or orientation. Protected information and complete reconstructive memory are not the same thing.

Mathematical work packages

TFET-1 — real-space carrier and invariant

Freeze a local Hamiltonian, coordinate operators, symmetry class, Fermi or mobility window and order of limits. Compute at least two equivalent or cross-checking invariants, preferably a spectral localizer and Bott/local marker. A Hall coefficient without a stable real-space index remains response geometry, not an admitted topological phase.

TFET-2 — disorder and scale tube

Vary positional disorder, coupling disorder, system size, boundary shape and detector resolution. Track the localizer gap, localization length, invariant, edge transport and rare-event tails separately. The topology must remain stable on a declared open parameter region rather than at one fitted sample.

TFET-3 — \(E_8\) chiral unit

Construct a source-native Hamiltonian, coupled-wire model or network whose low-energy edge realizes the \(E_8\) form and \(c_-=8\). Verify thermal or energy transport, absence of fractional bulk anyons, and robustness without inferring \(E_8\) from central charge alone.

TFET-4 — rank-24 collision experiment

Place \(E_8^3\) and Leech/Niemeier boundary candidates behind the same coarse bulk detector. Add current spectra, norm-two fields, defect-labelled traces and OPE data until the detector becomes injective on the declared family.

TFET-5 — orbifold and Monster enrichment

Gauge or coherently implement lattice inversion, retain the twisted sector and test the finite Monster witness stack already defined by the corpus: \(c=24\), \(V_1=0\), frame/Griess data, the \(2A\) split, marked \(23AB\), the order-46 junction and a nontrivial sewing relation.

TFET-6 — return and reconstruction

Require one carrier and action to reproduce the bulk index, edge response, defect operations and held-out records. Determine exactly which information is lost at each quotient. A catalogue replay or programmed scalar trace is not a physical reconstruction.

First computational and experimental targets

The first finite seed is already executable. An \(8\times8\) real-space Qi--Wu--Zhang carrier at mass \(-1\) is subjected to one frozen onsite-disorder field at strengths \(0,0.5,1,2\). Translation symmetry is broken at every nonzero strength, while the occupied-projector Bott index remains nontrivial and the finite spectral gap remains open. This establishes the computation path. A second certificate displaces the point carrier itself, recomputes every bond direction and cross-checks the periodic Bott index against an open-boundary spectral-localizer index at jitter strengths \(0,0.10,0.25,0.40\). Both retain one nontrivial class and both gaps remain open. This is structural disorder on a fixed locality graph, not yet a fully random graph.

The first exceptional layer is exact first at effective-edge level and now at continuum-local coupled-wire level. The E8 Cartan form defines an eight-channel bosonic invertible edge; arbitrary positive spatially varying interaction matrices preserve co-propagation and \(c_-=8\). A seven-wire irregular array with 48 mutually null interactions derives opposite E8 boundaries, while an eight-simple-current training record predicts the complete held-out 240-root shell. The same null schedule now has an exact two-local finite-qudit regularization with a width-independent bulk gap above its free-edge ground space. Clock-to-rotor convergence, continuum E8 anomaly recovery, a material-native carrier and random connectivity remain open. The first of those gaps is now narrowed: one compact clock mode converges to the fixed- coupling rotor with (1.11\times10^{-4}) relative gap error by (q=129). The next join is now typed more carefully. The topological (K{E8}) form does not select the positive constitutive energy matrix. Using (K^{-1}) as a charging kernel and using it as a capacitance/Gram matrix predict different eight-mode Coxeter fingerprints. Weak eight-port tomography must select and calibrate the physical branch before a nonlinear joined gap is meaningful. The subsequent sparse calculation is now compiled: unimodularity gives both branches one E8-theta kinetic-shell basis, while their distinct cosine step sets define different sixteen-neighbor hopping graphs. That solver has now resolved four projected levels through 117,361 states for both graphs. Their nonlinear spectra separate, but their first gaps still drift by 6.35% and 21.46% from cutoff four to six, so neither continuum gap is promoted. A restarted block solver now reaches 340,321 states; the next gap drifts shrink to 3.11% and 7.80% but still fail the same one-percent gate. The inverse branch also fails an earlier operator gate: its dense cosine steps cost up to 15 kinetic units, so shells below 15 omit at least one complete vacuum drive channel. Its first operator-complete basis has 3,721,681 states. The eligible (K)-kernel branch now reaches 794,161 states with a 0.295% gap increment and boundary probability below (1.9\times10^{-4}). It needs one confirming shell. Its odd first excitation is sine-bright and cosine-dark, so detector quadrature is part of the promotion gate. Shell 12 now supplies that confirmation: the provisional unit-coupling compact- cell gap is (0.1019267219), while all bulk, anomaly and material promotions remain open. Five ratios (u/U=0.25,0.375,0.5,0.75,1) now pass the same convergence and detector gates. They form a sampled stability spine; explicit channel and metric perturbations remain necessary for an open tube.

The material route is now typed rather than associative. A quasicrystal may supply a phason-bearing, translation-free backplane for the programmed E8 network, but icosahedral diffraction and higher-dimensional cut-and-project geometry do not establish E8. A blinded 48-device factorial design compares periodic, quasicrystalline and amorphous backplanes under identity and E8 relation programmes, then promotes only through gap, edge and held-out-current records.

At rank 24, the first blind detector has now been identified exactly. Thermal and bulk data collide, but the weight-one current dimensions are 744, 24 and 0 for E8-cubed, the Leech lattice theory and Moonshine. Only the zero-current branch proceeds to the existing Griess, marked-defect and sewing witness stack.

1. Apply a spectral-localizer/Bott pipeline to a controlled crystalline-to- amorphous Hamiltonian family and record where response, real-space index and mobility gap separate.
2. Reconstruct the continuum constitutive matrix by eight-port tomography, then test cutoff-stable nonlinear gap retention and chiral response on that measured branch.
3. Run the rank-24 blind-detector problem: give an inference process only bulk response first, then current, defect and OPE channels, and measure the first channel that separates \(E_8^3\), a rooted Niemeier theory and Leech.
4. Test whether an inversion-orbifold edge flows toward the same infrared witness under independently sampled local disorder or instead retains only the coarse \(c=24\) class.
5. Fabricate the periodic/quasicrystalline/amorphous backplane matrix and test whether the material changes the E8 programme beyond either main effect.

Promotion and kill gates

The arm advances only when:

• a real-space invariant is defined on the actual carrier;
• a spectral or mobility gap and locality bound protect it;
• boundary transport matches the bulk index;
• exceptional labels are derived from finer observables, not attached by dimension or central charge;
• the detector distinguishes smaller rival edge theories;
• disorder, finite size, calibration and provenance are carried through;
• one held-out record is predicted without refitting.

The direct exceptional claim is killed if every accessible record factors through the coarse bulk class or a smaller current/defect algebra. The translation-free topology result remains valuable after that null.

Corpus joins

• Exceptional horizon
• Disorder, cost, scale and rare-event docket
• Terminal return class
• Leech edge orbifold
• Monster proof certificate
• Pure-topological memory no-go
• Typed Monster reconstruction gate
• Quasicrystal backplane bridge
• E8 constitutive-metric gate
• E8 adaptive-shell nonlinear solver
• E8 joined-rotor nonlinear reconnaissance
• E8 restarted nonlinear eigensolver
• E8 operator-completeness gate
• E8 Cartan-kernel shell-10 spectroscopy
• E8 Cartan-kernel compact-cell gap
• E8 Cartan-kernel coupling-ratio spine
• E8 unequal-channel amplitude perturbation
• E8 channel-amplitude stability stencil
• E8 positive kinetic-metric perturbation
• E8 mixed perturbation corner
• E8 projected two-cell coupling
• E8 two-cell local-root audit
• E8 projected-chain width/root grid
• E8 projected-chain boundary comparison
• E8 local operator dictionary

Primary external anchors

• Agarwala and Shenoy, Topological Insulators in Amorphous Systems.
• Mitchell et al., Amorphous topological insulators constructed from random point sets.
• Zhao et al., Giant berry curvature in amorphous ferromagnet Co2MnGa.
• Lim, Mulligan and Teo, Partial fillings of the bosonic E8 quantum Hall state.
• Lam, Yamada and Yamauchi, Vertex operator algebras, extended E8 diagram, and McKay's observation on the Monster simple group.
• Makkar, Vertex algebras, topological defects, and Moonshine.
• Else et al., Quantum Many-Body Topology of Quasicrystals.