QH4 Pattern Matching
★ Open Conjecture

O(1) structural classification of ring positions by pattern type — the conjecture to derive — all priors established in Chapter 15 · Research Seed

Chapter Status

This chapter is a research seed. The prior results (Chapters 13–15, including all 9 theorems of the Pattern Mining Machine) are established and locked. The Pattern Matching conjecture stated here is not yet derived. This chapter records the conjecture precisely so future work begins from known ground, not guesswork.

Abstract

From WHERE to WHAT

Chapter 15 established the O(1) inverse formula: given position p, compute its Λ(ρ,γ,σ,θ) address in constant time. This tells us where position p is in the ring. The Pattern Matching conjecture asks a different question: what is position p — its structural role, its relationship to neighbours, and how to recognize a known pattern without searching.

Given a set of positions S ⊆ Z₂₅₆, classify S by its structural signature — without iterating through bins. The answer is O(1) per position, derived purely from (γ, σ, θ) arithmetic.


01 · Established Priors

Do Not Re-Derive — Reference Only

All of the following are proven and locked in Chapters 13–15. They are the starting point for the conjecture, not the subject of it.

PriorContentStatus
A1 — Ring constants256=4⁴, 4 vacuums, 252 active, 36 gates × 7 steps, 9 gates/epochLocked
A2 — Forward formulapos(γ,σ) = (7×(7(γ−1)+σ+⌊(γ−1)/9⌋)) mod 256252/252
A3 — Inverse formulalocate(p): m=(p×183)%256, adj correction, γ=div, σ=mod — O(1)252/252
A4 — Quadrant midpoints{32,96,160,224} — slot σ=4, equidistant at spacing 64Locked
T3 — Anchor conditionp is anchor ⟺ p mod 32 = 0 — O(1), no locate() requiredProven
T4 — Vacuum adjacencyp mod 64 ∈ {1,63} — O(1), no locate() requiredProven
J2 — Pattern classifier6-type classification in O(n) for sets of positionsProven

02 · The Conjecture

Structural Classification from Arithmetic

The six pattern types defined in J2 (Ch15) classify any set S by iterating locate() over each element. The Pattern Matching conjecture claims something stronger: given positions p₁, p₂, …, pₙ, their pattern type is determined O(1) from (γ, σ, θ) arithmetic — without calling locate() at all for individual comparisons.

Conjecture (to be derived)
Given positions p₁, p₂, ..., pₙ ∈ QH4 active ring,
their pattern type is determined O(1) by:
1. Compute locate(pᵢ) → (γᵢ, σᵢ, θᵢ) for each pᵢ
2. Compute epoch(γᵢ) = ⌊(γᵢ−1)/9⌋ + 1
3. Classify by which of {γ, σ, epoch, dist_to_vac} are equal
No search. No table. Pure arithmetic on (γ, σ) signatures.
Pattern NameDefinition
Bin clusterAll positions share the same gate γ — differ by multiples of 7
Step columnAll positions share the same step σ across different gates
Epoch stripeAll positions within one epoch (quarter of ring)
Vacuum-adjacentPosition is first or last slot of an epoch (1 step from vacuum)
MidpointPosition is slot σ=4 of its gate (centre of gate window)
Cross-epochSet of positions spanning multiple epochs

03 · Starting Questions

Five Questions to Open the Derivation

#Question
Q1What is the algebraic condition for two positions to be in the same gate? Is it (p₁×183 − p₂×183) mod 7 = 0 after vacuum correction?
Q2What is the condition for same step across gates? Does pᵢ mod 7 = constant hold directly?
Q3Can we detect a vacuum-adjacent position directly from p without locate()? (T4 does this — but can we also classify the pattern type of a set without locate()?)
Q4Does the distance between two positions in (γ,σ) space encode physical meaning in MPRC?
Q5What is the pattern signature of the 4 quadrant midpoints {32,96,160,224}?

Q1 and Q2 are the algebraic core. If the same-gate and same-step conditions can be read directly from p (without going through locate()), the full classifier becomes a sequence of modular arithmetic tests on raw position values — no address computation needed per comparison.


04 · MPRC Context

Physical Meaning of Pattern Type

In MPRC, the ring position p encodes a physical state: φ (position on the oloid), u-spinor direction, and quadrant polarity. Pattern types are not abstract classifications — they correspond to physical groupings:

Pattern TypeMPRC Interpretation
Gate clusterSame coherent u-axis window — same angular momentum range
Step columnSame phase position within each window — same φ modular offset
Epoch stripeSame quadrant — same polarity domain (Q⁰..Q³)
Vacuum-adjacentBoundary state — 1 step from a frame crossing at {0,64,128,192}
Midpointσ=4 structural node — centre of coherent window

Pattern matching in MPRC context means recognizing which physical state a position belongs to without scanning the ring — directly from the position value. This enables O(1) state identification across any modality.