← SCPN-Quantum-Control

Flagship result

Hardware observation · ibm_kingston (Heron r2) · April 2026

Backend-sensitive hardware observation of parity-sector and excitation-number correlated decoherence asymmetry in the XY Hamiltonian

The dynamical Lie algebra of the XY Hamiltonian on $n$ qubits splits under the parity operator into two equal blocks. In a 342-circuit campaign on IBM ibm_kingston, the parity-labelled sectors show a measurable leakage asymmetry under decoherence — peak $+17.5\%$ at depth 6, combined $p \ll 10^{-16}$. The paper is deliberately conservative about mechanism: the promoted claim is parity-sector and excitation-number correlated, not DLA-parity alone, and a second-backend control finds mixed signs. The full dataset, statistics, and an in-browser recompute of the headline number are on the results page.

Manuscripts

Synchronisation study · ibm_fez (Heron r2, 156q) · February 2026

Quantum simulation of coupled-oscillator synchronisation on a 156-qubit superconducting processor

A Kuramoto–XY synchronisation study with heterogeneous natural frequencies, simulator baselines, and ledgered IBM Heron r2 evidence. A Rust/PyO3-assisted pipeline computes entanglement entropy, Krylov complexity, OTOC scrambling, and Floquet discrete-time-crystal diagnostics for 2–16 qubits; the finite-size diagnostics are consistent with Berezinskii–Kosterlitz–Thouless scaling in the small systems studied. Legacy hardware rows are cited by committed artefact path rather than as aggregate validation.

Methods & software

A Rust-accelerated, topology-informed workflow for Kuramoto–XY quantum simulation and NISQ benchmarking

The software and experimental workflow behind scpn-quantum-control: mapping heterogeneous Kuramoto-type oscillator networks to XY spin Hamiltonians, compiling topology-informed variational ansätze, benchmarking classical and quantum simulation paths, and running reproducible NISQ hardware experiments. Python/Qiskit orchestration is combined with a Rust/PyO3 acceleration layer for hot-path kernels. The paper does not promote the topology-informed ansatz as hardware-general superior — it characterises where its inductive bias helps and where it does not.

Model + hardware test falsification

Fisher-information-inspired collective feedback in Kuramoto–XY Hamiltonians: exact small-system structure and a hardware falsification

A collective Fisher-information-inspired magnetisation-feedback term is added to the heterogeneous XY Hamiltonian: a diagonal term proportional to $-\lambda\,M^2/n$ that shifts magnetisation-sector energies without breaking the magnetisation conservation of the ideal XY model. Exact diagonalisation on $n = 4, 6, 8$ characterises the small-system structure, and the hardware test is reported as a falsification rather than a protection claim — the boundary is preserved honestly.

Mechanism-separation check · ibm_marrakesh (Heron r2)

Reduced-Pauli entanglement checks for DLA-sector leakage mechanisms on IBM Heron hardware

Parity-sector leakage asymmetry is a useful observable only if its mechanism can be separated from layout, prepared-state, and readout artefacts. This is a cost-bounded reduced-Pauli entanglement/tomography check for the promoted DLA-parity and Fisher-information-modified circuit families: 162 main circuits on ibm_marrakesh with physical qubits $[1,2,3,4]$, 2048 shots per main circuit. It is closer to targeted local-observable estimation than to full tomography, and reports where the resulting deviations from exact references are largest.

Control experiment · ibm_kingston dynamic circuits negative result

Monitored feedback versus open-loop control for Kuramoto–XY synchronisation on IBM dynamic circuits

Dynamic circuits allow monitored quantum-control policies to be tested on gate-model hardware. This preregistered four-qubit experiment on ibm_kingston pairs a monitored-feedback arm against a matched open-loop control across three dynamic rounds. The preregistered binary synchronisation endpoint is negative: the feedback arm does not improve on the matched control. The scientific value is methodological — a preregistered, backend-checked control-boundary result, reported as such rather than promoted into a controller claim.

Citing & archives

Every manuscript above carries its own Zenodo DOI (linked on each entry). The software framework itself is archived at Zenodo 10.5281/zenodo.18821929, and the framework-level SCPN Master Publications at Zenodo 10.5281/zenodo.17419678. Source code is under AGPL-3.0 at github.com/anulum/scpn-quantum-control. Author: Miroslav Šotek (ORCID 0009-0009-3560-0851).