Projected lattice theory and the dangerous cluster

Read the theorem labels as a dependency structure

The projected synthesis organizes several distinct steps. It begins with a finite lattice model, introduces a candidate four-dimensional sector, compares its Hamiltonian with an auxiliary mixed operator, reduces the remaining low-window problem to a band-effective model, and then transfers the result toward a continuum observable theory.

The source explicitly distinguishes proved abstract implications from theorem targets in the concrete lattice realization. Its master endpoint assumes the projected-sector and coercivity packages, thermodynamic and continuum comparison maps, isolated-band stability, observable continuity, finite-band localization and curvature approximation.

Thus a downstream theorem can be mathematically valid while its lattice hypotheses remain unproved. The useful question is which exact input produces each claimed output. Accepting the SMM/DTT interpretation does not discharge an analytic assumption.

The later vacuum-obstruction and auxiliary-framework manuscripts sharpen the front end. In particular, a sharp mixed-energy window must not be assumed to preserve the vacuum-orthogonal space. This guide uses their explicit puncture before any strict coercivity or Feshbach estimate.

Source: Projected lattice gauge synthesis, §§1.5–1.9, pp.7–10.

Construct the operator on the right state space

The mixed plaquette identity is an equality of functions of link matrices. After axial gauge, the associated finite matrix-sequence difference has a holonomy-twisted Fourier spectrum. The actual multiplication operator on wavefunctions acts by Kψ(X)=K(X)ψ(X). The physical Hamiltonian also contains derivatives with respect to link variables.

These are three different operators. The focused vacuum paper states this distinction explicitly. A direct sum of finite fiber spectra does not automatically give the quantum spectrum of the multiplication operator or of H.

A low multiplication window is generally infinite dimensional: if {K<ε} contains an open set of positive measure, infinitely many independent square-integrable functions can be supported there. The fact that the auxiliary circle has only N sites does not change that fact.

The projected programme therefore needs a concrete common-space realization. Its form comparison must specify the observable core, the operator represented there and the relation of that representation to the exact Hamiltonian. Finite packet approximations to holonomy data also need an approximation map and controlled error before their finite dimension can be used in an exact spectral argument.

Source: Projected lattice gauge synthesis, §§2.4 and 3.1–3.2, pp.13–20.

What the high-energy comparison would prove

After making the vacuum-orthogonal sector and stable form domain legitimate, suppose the Hamiltonian and mixed forms satisfy q_H[ψ]≥αq_K[ψ]−β∥ψ∥², with α>0 and β≥0. Let P_ε=1_[0,ε](K), Q_ε=I−P_ε on that sector.

The spectral theorem gives q_K[ψ]≥ε∥ψ∥² for vectors in the high range. Hence

The displayed regime makes the high compression of H−δ strictly positive. This is the input required to form its Hamiltonian Schur complement. The comparison itself is a theorem target in the model, including uniformity of α and β with respect to physical volume.

The β term matters. A positive auxiliary eigenvalue below β/α is not controlled by this inequality. Those modes remain in the low-window problem. Nor does the Hamiltonian compression estimate automatically imply a high transfer contraction when the window is not invariant under H.

For a literal multiplication window that splits Ω, use the corrected transfer formulation and puncture instead of assuming a stable excited-space decomposition.

Source: Projected lattice gauge synthesis, §§3.1–3.3, pp.15–20.

Use a model to isolate the dangerous cluster

Assume the low Feshbach form has a subspace E identified by a unitary W with a compact-resolvent model L. Its compression M is bounded below by W†LW−η_band, its complementary block R obeys R≥r_outI>0, and the coupling C has norm at most c_band.

Choose an energy cutoff E_* in L and pull its finite spectral projection back through W. This defines the dangerous cluster Π. Compact resolvent gives finite rank. The terminology means that only these low modes still threaten positivity; the complement can be made strictly positive by a Schur estimate.

This is an explicit sufficient remainder margin. It shows why the cutoff depends on both the model error and the coupling to the discarded part.

The existence of this comparison package is additional structure. It does not follow from counting a few low Fourier modes. The full low Hilbert space can be infinite dimensional while the compact-resolvent model isolates a finite set of potentially dangerous eigenvectors.

The exact fiber formula also quantifies the packet scale. A low eigenvalue λ≤ε requires |sin[(2πk+θ)/(2N)]|≤a_τ√ε/2. When the right side is below one, the phase must lie within 2N arcsin(a_τ√ε/2) of a multiple of 2πN. Near a zero branch, that width is approximately β_τ√ε. As the regulator changes, the relevant phase neighborhoods therefore change with the auxiliary circumference and the chosen threshold. A packet construction must track this scale, the number of packets, representation multiplicities and the error from replacing a varying phase by a representative. Merely selecting a finite set of representative holonomies does not prove that the omitted wavefunction dependence has finite rank or negligible energy.

Source: Projected lattice gauge synthesis, §§3.4–3.5, pp.21–23.

A finite model must exceed its error

Once the complement of Π is positive, eliminate it again. The cluster effective operator is F_Π^eff=M_Π−C_ΠR_Π⁻¹C_Π†. Positivity of this finite operator is equivalent to positivity of the low Feshbach operator under the preceding hypotheses.

Suppose a Hermitian model matrix L_Π approximates it with ∥F_Π^eff−L_Π∥≤η_eff. If L_Π≥(η_eff+μ)I for μ>0, then F_Π^eff≥μI. The finite error is deducted from the model’s smallest eigenvalue.

For a rank-one cluster this is a scalar inequality. For rank two, a convenient certificate L≥κI is

For example, the matrix with diagonal entries 0.4 and 0.3 and off-diagonal 0.1 has lowest eigenvalue about 0.238197. A certified operator error of 0.05 leaves a positive margin about 0.188197. That margin is a statement about the selected finite model, not yet a universal physical mass.

The final physical threshold remains the δ used when shifting H. A consistent proof tracks how its finite-model margin, projection errors and high-block bounds depend on every regulator parameter.

Source: Projected lattice gauge synthesis, §§3.5–3.6, pp.23–25.

Symmetry organizes multiplicities without creating a gap

A compact symmetry commuting with the dangerous-cluster operator lets one decompose the finite space into irreducible types and multiplicity spaces. Equivariant operators are constrained on each type and can be tested on the remaining multiplicities.

The projected synthesis uses the real commutant and real, complex and quaternionic representation types to describe compatible complex structures and oriented shadows. This can reduce the bookkeeping of finite matrices and show which mixings are permitted.

It does not determine the smallest eigenvalue of an allowed block. An arbitrary scalar multiple of the identity obeys every symmetry, including a scalar that tends to zero. Symmetry constraints and coercive estimates therefore play different roles.

The one-band and two-band cases are useful closed calculations because their model positivity and orientation data can be made explicit. Extending them to arbitrary observable tuples requires finite-band localization, control of resonances and bounds on all discarded off-block terms. Those packages remain visible hypotheses in the source rather than consequences of the labels “one band” or “two bands.”

Source: Projected lattice gauge synthesis, §§6–8, pp.33–46.

Carry the cluster across changing regulators

To compare different volumes or lattice spacings, place the operators in declared common comparison spaces. Strong convergence alone can lose finite-rank information. For example, rank-one projections onto basis vectors that drift to infinity converge strongly to zero, even though their ranks remain one.

A common isolated contour with norm-resolvent convergence gives stronger control. Integrating the resolvents yields

Once the right side is below one, the two finite-rank projections have equal rank. This is the mechanism behind stable transport of the dangerous cluster. The comparison maps, isolated contour and uniform resolvent bounds must be produced by the lattice analysis.

The vacuum must also converge strongly under these maps. Recovery sequences can then be orthogonalized by subtracting their vacuum component, allowing a uniform excitation-form bound to pass to the limit.

The source treats these as explicit producer-and-consumer interfaces: a concrete convergence package supplies the inputs, while an abstract theorem transfers the gap and cluster data. The transfer theorem does not construct the convergence package itself.

Source: Projected lattice gauge synthesis, §4, pp.25–28.

What remains before the Yang–Mills endpoint

A continuum projected theory must be compared through a separating algebra of renormalized Wilson-loop and bounded curvature observables. Equality on a dense, continuous and OS-cyclic core can identify complete vacuum representations, including their time semigroups and Hamiltonians.

The projected synthesis proposes a route from one-band and two-band sectors to arbitrary finite tuples, then from loop observables to bounded curvature generators. This route consumes resonance-isolation, tail, continuity and approximation bounds. A proof on selected toy bands does not automatically give equality on the full local observable core.

There is also a noncircularity condition. An independently constructed target can be matched by equality of correlations; a target defined merely to equal the projected limit supplies no identification theorem. The auxiliary framework therefore emphasizes intrinsic Yang–Mills properties, including its prescribed gauge group, field content, Ward identities and universality.

The projected construction is valuable as a detailed map of sufficient estimates and finite tests. Its master theorem remains conditional until those estimates are verified in the concrete regulator and the later vacuum and common-space corrections are incorporated.

Source: Projected lattice gauge synthesis, §§5 and 8–10, pp.28–33,41–51.

Continue reading

• Exact reduction and its domain conditions
• Real commutants and symmetry blocks
• The corrected vacuum puncture
• Comparison spaces and continuum identification