Comparative Coherence Decay of GHZ and Product States Under Idle Noise in IBM Heron Quantum Processors

Arturo Cerezo

Published Jan 16, 2026• 0 citations

Abstract

We report experimental measurements comparing the coherence decay of three-qubit Greenberger-Horne-Zeilinger (GHZ) entangled states versus locally coherent product states (|+++⟩) under idle time evolution on IBM Heron quantum processors. Across three different processor generations (Heron r1, r2, r3), we consistently observe that the GHZ state maintains higher fidelity than the product state as idle delay increases. Decay rate analysis reveals that |+++⟩ decays 2.4–13.8× faster than GHZ depending on the backend. A control experiment with |000⟩ confirms the decay mechanism is dominated by dephasing (T₂-type noise) rather than energy relaxation (T₁) or readout errors. These results demonstrate a structural dependence of coherence stability on the topology of the quantum state. Backends: ibm_fez (Heron r2), ibm_torino (Heron r1), ibm_boston (Heron r3)Configuration: 3 qubits, 4000 shots, 20 repetitions per delay pointDelays: 0, 500, 1000, 2000, 4000 dt

DOI: 10.5281/zenodo.18274446

Related papers

Comparative Coherence Decay of GHZ, W, and Product States Under Idle Noise in IBM Heron Quantum Processors: Extended Analysis with Dynamical Decoupling

Arturo Cerezo

Version 2.1: Extended experimental analysis comparing coherence decay of GHZ, W, and product states on IBM Heron quantum processors (r1, r2, r3). Key findings:- GHZ states decay 2.4–13.8× slower than product states across three processor generations- Dynamical decoupling with virtual Z-pulses (DD-Z) provides marginal improvement (+0.6%), while physical X-pulses (DD-X) degrade coherence by -24.4%- W-state performance is architecture-dependent: superior on r1/r3, inferior on r2- Weak non-Markovian noise signatures detected at 5 µs delay window This version extends v1.0 (DOI: 10.5281/zenodo.18274447) with W-state analysis, comprehensive DD protocol testing, and the DD-Z vs DD-X discovery. Backends: ibm_torino (r1), ibm_fez (r2), ibm_boston (r3)Configuration: 3-5 qubits, 4000 shots, delays 0-100,000 dt

Jan 17, 20260 citations

Observable-Dependent Population Retention in GHZ States Under Dephasing-Dominated Noise

Arturo Cerezo

We report a state- and observable-dependent retention advantage in operational fidelity proxies under dephasing-dominated noise across multiple quantum platforms. We do not claim enhanced off-diagonal GHZ coherence; rather, we compare measured observables with different channel sensitivities under identical physical noise: GHZ population fidelity Fpop = P(|00…0⟩) + P(|11…1⟩) (diagonal in the computational basis) versus an X-basis product-state control implemented as H⊗N → delay → H⊗N → measure. On IonQ Forte (5–20 qubits), parity visibility follows V(N) = 1.041 exp(−0.030N) with χ²/dof = 0.86 across 5–15 qubits. At 20 qubits, V departs by −8.2σ, consistent with correlated-noise onset; model comparison (AIC) favors a piecewise description with regime separation. On IBM Heron (3-qubit GHZ experiments), the fidelity advantage ΔF reaches 69.7 percentage points at maximum idle delay, with the population survival ratio G = 12.5× from raw counts without error mitigation. Three platforms with average two-qubit gate fidelities in the 94–98% range do not show a detectable advantage within our dataset, consistent with a graded performance dependence on hardware fidelity. The effect is GHZ-specific within our data: W-states and 1D cluster states show reduced fidelities (−12% and −38% relative to product controls), with no indication of protection. All fidelities are from raw shot counts; no ZNE, PEC, or post-selection was applied. These results motivate regime-aware benchmarks that treat the measured observable as a first-class experimental variable.

Feb 15, 20260 citations

Structural Protection of GHZ States Across Superconducting and Trapped-Ion Quantum Processors

Arturo Cerezo

Version 3.1: Extended cross-platform validation of structural coherence protection in GHZ entangled states. KEY FINDINGS:- GHZ states show 2.4-13.8× slower decay (IBM Heron) and +5-9% fidelity advantage (IonQ Forte) vs product states- Structural protection is GHZ-specific: W-states (-12%) and Cluster states (-38%) underperform product states on IonQ- Non-Markovian memory (N1) detected in trapped ions: Dynamical decoupling with Z-pulses improves GHZ protection by +2.99%, 5× stronger than IBM's +0.6%- Scaling confirmed from 3 to 7 qubits with 100% consistency across all measurements PLATFORMS:- IBM Heron: ibm_torino (r1), ibm_fez (r2), ibm_boston (r3)- IonQ Forte 1 (trapped ions, 36 qubits) EXPERIMENT DATES: January 15-18, 2026TOTAL INVESTMENT: ~$2,350 USD (IBM + AWS/IonQ) This version extends v2.1 (DOI: 10.5281/zenodo.18284373) with IonQ cross-platform validation, entanglement-type dependence analysis, and non-Markovian memory detection via dynamical decoupling.

Jan 18, 20260 citations

Decoherence of up to 8-qubit entangled states in a 16-qubit superconducting quantum processor

Asier Ozaeta, Peter L. McMahon

Abstract We report on the coherence of Greenberger–Horne–Zeilinger (GHZ) states comprised of up to 8 qubits in the IBM ibmqx5 16-qubit quantum processor. In particular, we evaluate the coherence of GHZ states with N = 1, …, 8 qubits 4 , as a function of a delay time between state creation and measurement. We find that the decay in coherence occurs at a rate that is linear in the number of qubits. This is consistent with a model in which the dominant noise affecting the system is uncorrelated across qubits.

arXiv:1712.07080Quantum Science and TechnologyMar 27, 201916 citations

Non-monotonic Fidelity Phase Diagram of GHZ States from N=2 to N=12 on IBM Quantum Hardware

Takayuki Takagi

Overview Systematic experimental characterization of Greenberger-Horne-Zeilinger (GHZ) states from N=2 to N=12 qubits on IBM Quantum's ibm_torino processor (133 qubits, Heron r2 architecture). Key Discovery The fidelity landscape exhibits unexpected non-monotonic behavior — contradicting the conventional expectation of monotonic decay with increasing qubit number. Five Distinct Regions Identified Region N Fidelity Characteristic Peak 3 0.951 Optimal qubit chain Valley 4, 5 ~0.14 Decoherence collapse Recovery 6–8 ~0.69 V-shaped recovery Cliff 9 0.40 Stabilization 10–12 ~0.33 Fidelity floor Principal Results V-shaped recovery: Fidelity increases from F=0.07 (N=5) to F=0.73 (N=6), demonstrating that qubit chain quality dominates over circuit depth Mermin violation: M₃ = 3.65 ± 0.01, exceeding the classical bound by 143 standard deviations Asymmetric decoherence: For N≥9, the |1⟩⊗N component vanishes entirely while |0⟩⊗N maintains ~30% probability Implications This 11-point dataset suggests that "fidelity phase diagrams" could serve as a new tool for quantum processor characterization, complementing existing benchmarks. Reproducibility All IBM Quantum Job IDs are provided for independent verification.

Dec 4, 20251 citation

Structural Protection of GHZ States: Definitive Cross-Platform Validation Across Five Quantum Processors (v4.0)

Arturo Cerezo

Version 4.0: Definitive experimental validation of structural coherence protection (N2) in GHZ entangled states across five quantum computing platforms. KEY RESULTS:- IonQ Forte: +7.27% GHZ advantage over product states (p = 7.17×10⁻²⁷, Cohen's d = 6.11)- IBM Heron: 2.4–13.8× slower decay for GHZ vs product states- Rigetti Ankaa-3, IQM Garnet, IQM Emerald: No N2 detected (negative controls) MAIN FINDINGS:1. N2 is real and statistically incontrovertible (p << 5σ particle physics standard)2. N2 is cross-platform: confirmed on superconducting (IBM) and trapped-ion (IonQ) technologies3. N2 requires high-fidelity hardware: >99.5% 2Q gate fidelity threshold4. N2 is GHZ-specific: W-states and Cluster states show no protection5. Negative results on Rigetti/IQM provide mechanistic insight into fidelity thresholds PLATFORMS TESTED:- IBM Heron (ibm_torino, ibm_fez, ibm_boston)- IonQ Forte 1 (trapped ions, 36 qubits)- Rigetti Ankaa-3 (84 qubits)- IQM Garnet (20 qubits)- IQM Emerald (5 qubits) Total experimental investment: ~$3,700 USD This version incorporates Panel Cerezo methodology recommendations and extends v3.1 (DOI: 10.5281/zenodo.18291164) with definitive statistical analysis meeting all rigor criteria. Experiment dates: January 15-19, 2026

Jan 19, 20260 citations