Parallel Extension Of Binary Circuit Gives Threshold-Lift Cut Rank¶
Claim/Theorem¶
Let M be a binary matroid on a ground set E=P\sqcup Q, where P and Q are both nonempty. Assume:
Pis a parallel class ofM;- if
pis any representative of that parallel class in the simplification\operatorname{si}(M), thenQ\cup\{p\}is a circuit of\operatorname{si}(M).
Write
For a cut L\subseteq E, set
Then the connectivity function of M depends only on (p_L,q_L) and is given by
Equivalently,
Hence for any full-row-rank binary stabilizer matrix H whose column matroid is M, the stabilizer cut-rank function satisfies
and is ordinarily hidden-vertex graph-cut representable by the same four-hidden-bit quadratic submodular energy from [[parallel-class-affine-basis-family-is-hidden-vertex-graph-cut-representable.md]].
Conversely, every family H_{m,s} from [[parallel-class-affine-basis-family-is-hidden-vertex-graph-cut-representable.md]] has this matroid structure. More strongly, every binary matroid satisfying the above hypotheses is isomorphic to one of those H_{m,s} families. So the threshold-lift pattern has a clean presentation-invariant characterization: it is exactly the class of binary matroids obtained by replacing one element of a circuit by a parallel class.
This is a derived structural theorem.
- By [[cross-cut-stabilizer-rank-rank-formula.md]], any stabilizer cut-rank function from a binary matrix
Hequals the connectivity function of its column matroid, so it is enough to compute\lambda_M. - Because
Q\cup\{p\}is a circuit in\operatorname{si}(M), the setQis independent of ranks, andp\in\operatorname{cl}(Q)butp\notin\operatorname{cl}(R)for every proper subsetR\subsetneq Q. - Therefore a subset with parameters
(p_L,q_L)has rank
Indeed:
- a nonempty subset of the parallel class P has rank 1;
- any proper subset of Q is independent and does not span p;
- the full set Q spans p because Q\cup\{p\} is a circuit.
4. Applying the same formula to the complement (m-p_L,s-q_L) gives
which simplifies exactly to the threshold-lift formula above.
5. The hidden-vertex realization now follows immediately from [[parallel-class-affine-basis-family-is-hidden-vertex-graph-cut-representable.md]], because the right-hand side depends only on the pair (m,s) and the counts (p_L,q_L).
6. For the converse structural characterization, choose any q_1\in Q. Since Q\cup\{p\} is a circuit, the set
is independent of size s, hence a basis of the represented vector space. In coordinates with respect to B,
p=e_1,- the
s-1elements ofQ\setminus\{q_1\}becomee_2,\dots,e_s, - the remaining element
q_1becomese_1+\cdots+e_s,
because in a binary circuit the unique dependence is the total sum of the circuit elements.
Thus every such matroid is representable in exactly the normal form used in [[parallel-class-affine-basis-family-is-hidden-vertex-graph-cut-representable.md]].
Consequences for the current frontier:
- the threshold-lift realization is not tied to one chosen generator matrix; it is a presentation-invariant binary-matroid phenomenon;
- the current positive result extends exactly to the structural class “parallel extension of a circuit,” not just one hand-picked
H_{m,s}coordinate family; - this is a wider theorem in invariant form, but not yet a larger isomorphism class than the explicit family already on disk;
- any genuinely new positive theorem must therefore move beyond circuit-plus-parallel-class structure, while any negative theorem must find a different witness outside this structural class.
Dependencies¶
- [[cross-cut-stabilizer-rank-rank-formula.md]]
- [[parallel-class-affine-basis-family-is-hidden-vertex-graph-cut-representable.md]]
Conflicts/Gaps¶
- This node characterizes the present threshold-lift family invariantly, but it does not classify all hidden-vertex representable stabilizer cut-rank functions.
- The theorem is derived from basic binary-matroid circuit properties plus the explicit hidden-vertex construction already on disk; it is not stated verbatim in the cited papers.
- Even this invariant structural theorem remains at the level of auxiliary-variable graph-cut expressibility and does not yet yield routing-style
CD(T_n,G)semantics on the physical qubit set.
Sources¶
10.48550/arXiv.2109.1459910.1016/j.dam.2009.07.001