Transjunction coupling — ON vs OFF

Matched ablation on a T03 cluster slice: 3×26Q = 78Q, same seeds and 1024 shots, coupled transjunctions enabled vs disabled. Primary metric: localized inter-context correlation (ICC) on ring-linked qubit pairs, with QPC readout mitigation (QPC noise-reduction module).

Plain English

What this test is: A matched ON vs OFF experiment: same 78Q circuit, transjunction coupling enabled vs disabled — only the bridge wiring changes.
Why we did it: ICC asks whether coupling matters. This ablation isolates transjunctions on designated bridge qubit pairs with readout mitigation applied.
What QPC computed: Localized ICC on ring-linked pairs · 1024 shots · T03 cluster slice · QPC noise-reduction module on readout · Fez and Marrakesh backends.
Headline result: Both Heron backends pass — |Δ localized ICC| > 0.05 (mitigated): Fez +0.064 · Marrakesh +0.069 on bridge pairs.
What we claim — and do not: Claim: turning transjunctions ON raises localized ICC on bridged pairs. Do not claim: full-width ICC across all 26 qubits (expected ~0 — most pairs were never bridged).

Verdict

Both IBM Heron backends pass |Δ localized ICC| > 0.05 (readout mitigated, June 2026).

Coupled runs show higher localized ICC than decoupled — consistent with transjunction imprint on designated coupling pairs.

Full-width ICC (all 26 aligned qubit indices) stays near zero (~0.003 Δ) — expected: most pairs were never bridged. This page reports designated coupling pairs only.

Results by backend

Backend	Coupled ICC (mit)	Decoupled ICC (mit)	Δ ICC	Δ localized ZZ	Coupled job	Decoupled job
ibm_fez	0.072	0.008	+0.064	+0.117	`d8fh39ralsvc7391oe30`	`d8fh3ro7jphs739m98n0`
ibm_marrakesh	0.095	0.026	+0.069	+0.156	`d8fi333alsvc7391pvag`	`d8fi34jalsvc7391pve0`

Verify jobs in the IBM Quantum console. Machine-readable summary: available under partner NDA (job IDs above).

Proof figure — Marrakesh (primary)

Mean localized ICC (readout mitigated): coupled vs decoupled

Coupled ON

0.095

Coupled OFF

0.026

Δ = +0.069 · threshold 0.05 shown as reference line in analysis JSON

Method (short)

Layout: 3 contexts × 26Q morphogrammatic encoding (depth 8, seeds 7000–7002).
ON: coupled transjunctions (proprietary bridge wiring).
OFF: identical context circuits, bridges removed.
Observable: Bridge-local ICC on transjunction-linked qubit pairs (indices not published).
Noise layer: proprietary readout mitigation on bridge subset → recompute ICC/ZZ.

Developer notes (MD) Noise Reducer report

Honest scope

This is a controlled coupling ablation, not a domain outcome claim (see QQ pilot for prediction ablation).
Ring wrap pair (ctx2↔ctx0) can show weaker ICC than adjacent pairs — see per-pair JSON.
Complements open-instance scale evidence (T01–T05) on the same Heron fleet.