Skip to content

Lesson 07 - Intrinsic Matroid Frontier

Goal

Understand the intrinsic route: prove large balanced cut rank directly in the original qubit parity-check matroid. The current canonical target is dense tangle breadth.

Prerequisites And Diagnostic Checks

  1. What is matroid connectivity \(\lambda_M(L)\)?
  2. Why does \(\lambda_M(L)=\chi_L(\mathcal S)\) for the stabilizer support matrix?
  3. What does a balanced cut mean in the hardware-separator argument?
  4. What is a tangle, at the level of "a consistent orientation of low-order separations"?

Concrete Motivation

If classical routing semantics are fragile, one can instead try to prove the invariant lower bound directly:

\[ \chi_L(\mathcal S_n)=\Omega(n) \]

for every hardware-balanced cut \(L\).

That would be enough for the stabilizer cut-rank functional and the submodular CD theorem. The challenge is that generic good-code facts do not imply it. The proof must use family-specific structure of the original qubit matroid.

Worked Example Before Abstraction

A graph expander has no sparse balanced edge cut. The analogous matroid statement is not "many edges cross." It is:

every balanced partition has high rank connectivity.

For a stabilizer matrix \(H\), this means:

\[ \operatorname{rank}(H_L)+\operatorname{rank}(H_R)-\operatorname{rank}(H) \]

is linear for every balanced \(L\).

The problem is that local tester expansion or good distance can live in auxiliary structures without forcing this original-qubit rank expression to be large.

Canonical Target

The current graph compresses the intrinsic route into:

The target is:

\[ H_{\mathrm{dense}}(\beta): \quad \text{the original qubit parity-check matroid has a tangle of order }\Omega(n) \text{ and sufficiently dense breadth.} \]

If this holds, then every \(\beta\)-balanced cut has linear \(\chi_L\).

Three Intrinsic Macro-Routes

The graph currently says broad intrinsic work reduces to three family-specific lifts:

They are:

  1. grid domination to dense original-matroid concentration;
  2. robust flower template exclusion or concentration on balanced cuts;
  3. tester-side irreducibility to original-qubit matroid connectivity.

Current sources do not close any of these.

Supporting stop nodes include:

Conjecture Linkage

This route would close a strong invariant version of the Conjecture 3 barrier:

  1. prove dense original-matroid connectivity for the target Quantum Tanner family;
  2. conclude linear \(\chi_L\) on balanced cuts;
  3. feed \(\chi_L\) into the stabilizer cut-rank functional;
  4. get syndrome-depth lower bounds on low-separator hardware.

The open step is the first one.

What This Does And Does Not Prove

This proves:

  • dense tangle breadth is a clean sufficient intrinsic target;
  • several other intrinsic candidates are subordinate mechanisms toward that target;
  • generic currently sourced matroid/tangle/LTC theorems do not close the target.

This does not prove:

  • dense tangle breadth for Quantum Tanner codes;
  • that the intrinsic route is impossible;
  • that local tester expansion cannot help after a new family-specific lift is found.

Active Recall

  1. State the balanced cut-rank target in matrix language.
  2. Why is dense tangle breadth a sufficient target?
  3. Why are generic tangle theorems not enough?
  4. What are the three family-specific lift routes?
  5. Which route feels closest to the Quantum Tanner construction, and why?

Next-Step Handoff

Next lesson: zoom in on the construction-level local-block route, which is the most concrete way the Quantum Tanner structure might force balanced cut rank.