Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2603.29525 · Quantum Science and Technology

Non-perturbative CPMG scaling and qutrit-driven breakdown under compiled superconducting-qubit control: a single-qubit study

Jun Ye

quant-phSubmitted Mar 31, 2026• Announced Apr 1, 2026• 1 citation• Revised Apr 1, 2026 (v2)
v1 · Mar 31, 2026v2 · Apr 1, 2026v3 · Jun 15, 2026
arXivHTMLPDF

Abstract

Decoherence in superconducting qubits emerges from the interplay of multilevel dynamics and structured environmental noise, yet perturbative models cannot capture all resulting signatures. Here, EmuPlat couples instruction-set-architecture-level waveform generation to the hierarchical equations of motion (HEOM) under $1/f$ non-Markovian pure dephasing. In the resulting non-perturbative regime -- where filter-function predictions become quantitatively uninformative -- CPMG scaling of a three-level superconducting transmon yields one calibration result, two physical findings, and one structural null. Y-CPMG exhibits axis-dependent scaling-law breakdown -- non-monotonic decoherence, partial coherence revival, and pronounced X--Y population asymmetry ($0.204$ vs ${<}\,0.01$) -- driven by third-level anharmonicity amplified by bath memory; X-CPMG maintains well-behaved power-law scaling with a finite-$n$ transient excess consistent with non-Markovian bath-memory effects. The structural null is equally informative: waveform-level differences -- Standard versus VPPU realizations -- remain undetectable across all coupling strengths, establishing that rotating-frame pure-dephasing coupling renders control-layer detail invisible to scaling observables. These findings define testable predictions, the most experimentally accessible requiring only qualitative verification.

Comments
28 pages, 5 figures, 43 pages SI, comments/suggestions are welcomed

Related papers

Simulation of one and two qubit superconducting quantum gates in the presence of non-Markovian 1 / f noise

Yinjia Chen, Shuocang Zhang, Qiang Shi

Non-Markovian $1/f$ noise consists a dominant source of decoherence in superconducting qubits, yet its slow nature poses a significant challenge for accurate simulation. Here we develop a hierarchical equations of motion (HEOM) framework that enables efficient and reliable modeling of qubit dynamics and gate operations under $1/f$ noise. By using the approach, it is first shown that perturbative quantum master equations may fail to reproduce the correct dephasing dynamics of a qubit coupled to slow baths. We then analyze dynamical decoupling sequences by including effects of finite pulse duration. It is found that different pulse sequences results in different behavior in error accumulation: all X-CPMG sequences exhibit linear scaling with parity effects, Y-CPMG follows quadratic growth, and alternating XY-type sequences can suppress the error accumulation significantly. Finally, we extend the framework to two-qubit cross-resonance (CR) gates, reconstructing the full Choi matrix and Pauli Transfer Matrix (PTM) to identify the incoherent error induced by $1/f$ noise. Together, these results establish HEOM as a robust methodology for simulating the environmental noise in superconducting circuits and provide new insights into error mechanisms in both single- and two-qubit gates.

arXiv:2509.07693•Physical review. B./Physical review. B•Feb 10, 2026•3 citations

Engineering Non-Linear Decay Dynamics: Pulse-Level Control and Software-Defined Qubit Rescue on Superconducting Processors

K S, Unnikuttan

The scalability of Noisy Intermediate-Scale Quantum (NISQ) devices is currently constrained by material defects, specifically Two-Level Systems (TLS) that induce resonant decoherence in superconducting qubits. This study presents a comprehensive experimental analysis using the IBM Quantum ibm_fez processor to demonstrate "Software-Defined Hardware" optimization. By employing a novel "Instruction-Level Calibration Injection" technique, we bypass standard compiler constraints to inject continuous off-resonant AC Stark drives ($N_{shots} = 4096$). Methodology The experiment utilizes a Floquet engineering approach to perform pulse-level Hamiltonian engineering. We implement custom instruction-level calibrations to apply continuous off-resonant AC Stark drives, effectively modifying the qubit frequency to avoid resonant interactions with defect states. Key Findings The application of this protocol yielded three primary results: Spectral Decoupling: Successfully restored Ramsey fringes in a defect-limited qubit (Q150), achieving a coherence time of $T_2^* \approx 6.36 \: \mu\text{s}$. Protocol Benchmarking: The "Stark Rescue" protocol demonstrated a statistical advantage over standard CPMG dynamical decoupling. It extended coherence time by 11.3% ($T_2^{Stark} = 8.17 \: \mu\text{s}$ vs. $T_2^{CPMG} = 7.34 \: \mu\text{s}$; $p = 0.042$, Bootstrap t-test). Topological Simulation: Digital simulation of a topological domain wall under engineered correlated noise revealed a sigmoidal stability threshold ($SSE \approx 0.0008$), significantly outperforming standard exponential models ($SSE \approx 0.0085$). Conclusion These results confirm that pulse-level Hamiltonian engineering can effectively reclaim compromised hardware resources on current NISQ platforms. Keywords NISQ, Superconducting Qubits, Two-Level Systems (TLS), Floquet Engineering, AC Stark Effect, IBM Quantum, Dynamical Decoupling, Quantum Control.

•Dec 27, 2025•784 citations

Directed Recoherence in Structured Non-Markovian Environments: A Predictable Regime for Passive Coherence Enhancement

Ferreira, Manuel Rui Santos

Directed Recoherence in Structured Non-Markovian Environments presents a rigorously defined operating regime of open quantum system dynamics in which environmental memory leads not to oscillatory revivals, but to directed, monotonic recoherence toward a long-lived metastable coherence plateau. The work introduces a minimal and physically standard two-bath pure-dephasing model composed of a broadband Markovian noise bath and a narrowband non-Markovian reference bath. Within this framework, a dimensionless recoherence parameter R = λ_R² / [ γ_N Γ (1 + (δ/Γ)² ) ] is shown to govern a sharp transition between ordinary decoherence and a directed recoherence regime. When R exceeds a threshold of order unity, the system exhibits a single coherence minimum followed by monotonic recoherence and a metastable phase-locked plateau over experimentally relevant timescales. Exact closed-form solutions for the coherence dynamics are derived, alongside falsifiable predictions including bounded phase deviation relative to the reference bath. Numerical simulations confirm the existence of a critical threshold and distinguish this regime from known non-Markovian revivals, decoherence-free subspaces, driven steady states, and many-body prethermalization. The manuscript reframes Self-Initiated Phase Re-Coherence (SIPR) not as a new physical principle, but as a previously under-characterized regime within standard quantum open-system theory. Potential experimental realizations are discussed for superconducting qubits with Purcell-filtered environments and trapped ions with engineered sidebands. A companion conceptual document explores the epistemic and interpretational implications of measurement, observation, and environmental structure without modifying the physical results presented here.

•Dec 13, 2025•0 citations

Amplitude-Phase Coupling and Average-Hamiltonian Cancellation in Figure-8-Driven Transmon Qubits: Theory and Phase-Resolved Experiments on Two Devices

Kosaku Tabuchi

We report phase-resolved measurements of amplitude-phase coupling (APC) in strongly driven superconducting transmon qubits, a symmetry-based mechanism analysis, and follow-up experiments on a second device. On a Rigetti Ankaa-3 transmon (1000 shots per point, five drive amplitudes, matched root-mean-square drive), the fitted optimal control phase $\phi^{*}(A)$ drifts by more than 3 rad across the explored amplitude range for both a rectangular reference pulse and a self-intersecting Figure-8 trajectory, and the phase contrast of the two waveform families exhibits a reproducible low-/high-amplitude crossover. Theoretically, we prove that in the resonant rotating frame of the three-level Duffing model a constant-phase rectangular drive cannot produce any phase-dependent population response: $C_{Rect}=0$ exactly. Any measured rectangular phase contrast is therefore a probe of physics outside this model (beyond-rotating-wave carrier-phase dependence, control-chain distortion, readout imperfections). The Figure-8 waveform breaks the symmetry intrinsically through its 2:1 Fourier content; a third-order resonance $\omega+\omega-2\omega=0$ mediated by the anharmonic level $|2\rangle$ predicts $C_{F8}\propto A^{6}N^{2}$ at weak drive, confirmed by direct numerical propagation (fitted log-log slope $5.97\pm0.02$; loop-number ratio 2434 vs. predicted 2500). On the successor device Cepheus-1-108Q, the predicted phase contrast lies below the binomial shot-noise floor at the accessible amplitudes, but the same dataset reveals a stronger signature of the underlying first-order average-Hamiltonian cancellation: the phase-averaged survival probability of the Figure-8 train remains pinned at $\langle P(|0\rangle)\rangle_{\phi}\approx0.985$ across all tested conditions, while the isenergetic rectangular reference sweeps the full range 0.24-0.93—an approximately 70-fold suppression of the Rabi-drive response. Together these results identify APC as a measurable control-landscape feature with distinct intrinsic and extrinsic components, and demonstrate a calibration-free, amplitude-insensitive idling property of Figure-8 pulse trains.

•Jul 6, 2026•0 citations

HEOM-in-Calibration-Loop: Exposing Non-Markovian Bath Signatures That Markovian Calibration Elides in Superconducting-Qubit Tune-Up

Jun Ye

Closed-loop superconducting-qubit calibration has matured into DAG-orchestrated protocol chains, yet published frameworks treat the bath via a Markovian master equation or a phenomenological likelihood, absorbing bath structure into fit residuals instead of reporting it as a diagnostic. We integrate a QuTiP 5.x hierarchical-equations-of-motion (HEOM) solver driven by a Tier-1 1/f Burkard bath into a multi-protocol calibration DAG (Rabi -> {Ramsey || T1}) and benchmark it against sesolve and mesolve on a frozen platform in a pulse-level simulator (no hardware validation). The Ramsey channel carries the headline: the Markovian fit is censored by its exponential-family numerical ceiling, while HEOM recovers a physical revival envelope whose primary T2* separates from the Markovian reference by at least 13x at 95% independent-bootstrap confidence within the HEOM-feasible budget; the point-estimate ratio reaches >=28x on the 50-point primary-t1 grid and ~72x on the 30-point biexp-family tau_aw pivot at L=5. Rabi contrast falls 2.17% below mesolve on a noise-limited 30-point grid; the paired-bootstrap CI crosses zero, so this channel corroborates rather than independently establishes the non-Markovian signature. T1 decay shape matches across backends (beta=1.000), yet HEOM's initial occupation drops from 1.000 to 0.879 -- a bath-dressed contamination stable under a 16-point densification. The DAG adds 9.62 us average per-protocol scheduling overhead, no meaningful latency penalty at protocol granularity. HEOM-in-loop thereby changes what calibration reports: bath structure appears as a quantifiable residual rather than a hidden confound.

quant-pharXiv:2604.21458•Apr 24, 2026•0 citations

The Topological Decoherence Hypothesis: A Unified Origin for Anomalous Noise in Matter-Based Qubits

Daniel Edward Gruenberg

Every matter-based qubit platform—superconducting circuits, trapped ions, nitrogen- vacancy centres—exhibits decoherence in excess of what the standard two-level error model predicts from known noise sources. In superconducting qubits, two-level system (TLS) defect coupling accounts for 30–50% of the dephasing budget with debated mi- croscopic origin. In trapped ions, anomalous motional heating exceeds Johnson noise predictions by two to six orders of magnitude. In NV centres, residual pure dephasing persists after the nuclear spin bath and all identified sources are eliminated. These phenomena are platform-specific in detail but share a common structural feature: they all involve coupling of the qubit’s spin-projection state to spatial or motional degrees of freedom that the two-level model does not describe. Photonic qubits, by contrast, exhibit none of these anomalies. We propose that this asymmetry has a topological origin. A recent classifica- tion of electromagnetic field configurations into open (Class O, massless) and closed (Class C, massive) Poynting topologies [1] implies that massive particles carry internal structure—chirality, spatial mode geometry, Beltrami eigenstate classification—beyond their spin projection. Photons do not. The anomalous decoherence channels observed in every matter-based platform, and absent in every photonic platform, are the envi- ronmental coupling to this internal topological structure that the standard qubit model omits. This is not a small correction: the topological channels account for 5–50% of the observed decoherence, depending on the platform, and their platform-specific coupling fractions are consistent with known spin-to-spatial-mode coupling mechanisms at each scale.

•Mar 27, 2026•0 citations
Decoherence in superconducting qubits emerges from the interplay of multilevel dynamics and structured environmental noise, yet perturbative models cannot capture all resulting signatures. Here, EmuPlat couples instruction-set-architecture-level waveform generation to the hierarchical equations of motion (HEOM) under 1/f1/f1/f non-Markovian pure dephasing. In the resulting non-perturbative regime -- where filter-function predictions become quantitatively uninformative -- CPMG scaling of a three-level superconducting transmon yields one calibration result, two physical findings, and one structural null. Y-CPMG exhibits axis-dependent scaling-law breakdown -- non-monotonic decoherence, partial coherence revival, and pronounced X--Y population asymmetry (0.2040.2040.204 vs < 0.01{<}\,0.01<0.01) -- driven by third-level anharmonicity amplified by bath memory; X-CPMG maintains well-behaved power-law scaling with a finite-nnn transient excess consistent with non-Markovian bath-memory effects. The structural null is equally informative: waveform-level differences -- Standard versus VPPU realizations -- remain undetectable across all coupling strengths, establishing that rotating-frame pure-dephasing coupling renders control-layer detail invisible to scaling observables. These findings define testable predictions, the most experimentally accessible requiring only qualitative verification.
Non-Markovian 1/f1/f1/f noise consists a dominant source of decoherence in superconducting qubits, yet its slow nature poses a significant challenge for accurate simulation. Here we develop a hierarchical equations of motion (HEOM) framework that enables efficient and reliable modeling of qubit dynamics and gate operations under 1/f1/f1/f noise. By using the approach, it is first shown that perturbative quantum master equations may fail to reproduce the correct dephasing dynamics of a qubit coupled to slow baths. We then analyze dynamical decoupling sequences by including effects of finite pulse duration. It is found that different pulse sequences results in different behavior in error accumulation: all X-CPMG sequences exhibit linear scaling with parity effects, Y-CPMG follows quadratic growth, and alternating XY-type sequences can suppress the error accumulation significantly. Finally, we extend the framework to two-qubit cross-resonance (CR) gates, reconstructing the full Choi matrix and Pauli Transfer Matrix (PTM) to identify the incoherent error induced by 1/f1/f1/f noise. Together, these results establish HEOM as a robust methodology for simulating the environmental noise in superconducting circuits and provide new insights into error mechanisms in both single- and two-qubit gates.
The scalability of Noisy Intermediate-Scale Quantum (NISQ) devices is currently constrained by material defects, specifically Two-Level Systems (TLS) that induce resonant decoherence in superconducting qubits. This study presents a comprehensive experimental analysis using the IBM Quantum ibm_fez processor to demonstrate "Software-Defined Hardware" optimization. By employing a novel "Instruction-Level Calibration Injection" technique, we bypass standard compiler constraints to inject continuous off-resonant AC Stark drives (Nshots=4096N_{shots} = 4096Nshots​=4096). Methodology The experiment utilizes a Floquet engineering approach to perform pulse-level Hamiltonian engineering. We implement custom instruction-level calibrations to apply continuous off-resonant AC Stark drives, effectively modifying the qubit frequency to avoid resonant interactions with defect states. Key Findings The application of this protocol yielded three primary results: Spectral Decoupling: Successfully restored Ramsey fringes in a defect-limited qubit (Q150), achieving a coherence time of T2∗≈6.36 μsT_2^* \approx 6.36 \: \mu\text{s}T2∗​≈6.36μs. Protocol Benchmarking: The "Stark Rescue" protocol demonstrated a statistical advantage over standard CPMG dynamical decoupling. It extended coherence time by 11.3% (T2Stark=8.17 μsT_2^{Stark} = 8.17 \: \mu\text{s}T2Stark​=8.17μs vs. T2CPMG=7.34 μsT_2^{CPMG} = 7.34 \: \mu\text{s}T2CPMG​=7.34μs; p=0.042p = 0.042p=0.042, Bootstrap t-test). Topological Simulation: Digital simulation of a topological domain wall under engineered correlated noise revealed a sigmoidal stability threshold (SSE≈0.0008SSE \approx 0.0008SSE≈0.0008), significantly outperforming standard exponential models (SSE≈0.0085SSE \approx 0.0085SSE≈0.0085). Conclusion These results confirm that pulse-level Hamiltonian engineering can effectively reclaim compromised hardware resources on current NISQ platforms. Keywords NISQ, Superconducting Qubits, Two-Level Systems (TLS), Floquet Engineering, AC Stark Effect, IBM Quantum, Dynamical Decoupling, Quantum Control.
Directed Recoherence in Structured Non-Markovian Environments presents a rigorously defined operating regime of open quantum system dynamics in which environmental memory leads not to oscillatory revivals, but to directed, monotonic recoherence toward a long-lived metastable coherence plateau. The work introduces a minimal and physically standard two-bath pure-dephasing model composed of a broadband Markovian noise bath and a narrowband non-Markovian reference bath. Within this framework, a dimensionless recoherence parameter R = λ_R² / [ γ_N Γ (1 + (δ/Γ)² ) ] is shown to govern a sharp transition between ordinary decoherence and a directed recoherence regime. When R exceeds a threshold of order unity, the system exhibits a single coherence minimum followed by monotonic recoherence and a metastable phase-locked plateau over experimentally relevant timescales. Exact closed-form solutions for the coherence dynamics are derived, alongside falsifiable predictions including bounded phase deviation relative to the reference bath. Numerical simulations confirm the existence of a critical threshold and distinguish this regime from known non-Markovian revivals, decoherence-free subspaces, driven steady states, and many-body prethermalization. The manuscript reframes Self-Initiated Phase Re-Coherence (SIPR) not as a new physical principle, but as a previously under-characterized regime within standard quantum open-system theory. Potential experimental realizations are discussed for superconducting qubits with Purcell-filtered environments and trapped ions with engineered sidebands. A companion conceptual document explores the epistemic and interpretational implications of measurement, observation, and environmental structure without modifying the physical results presented here.
We report phase-resolved measurements of amplitude-phase coupling (APC) in strongly driven superconducting transmon qubits, a symmetry-based mechanism analysis, and follow-up experiments on a second device. On a Rigetti Ankaa-3 transmon (1000 shots per point, five drive amplitudes, matched root-mean-square drive), the fitted optimal control phase ϕ∗(A)\phi^{*}(A)ϕ∗(A) drifts by more than 3 rad across the explored amplitude range for both a rectangular reference pulse and a self-intersecting Figure-8 trajectory, and the phase contrast of the two waveform families exhibits a reproducible low-/high-amplitude crossover. Theoretically, we prove that in the resonant rotating frame of the three-level Duffing model a constant-phase rectangular drive cannot produce any phase-dependent population response: CRect=0C_{Rect}=0CRect​=0 exactly. Any measured rectangular phase contrast is therefore a probe of physics outside this model (beyond-rotating-wave carrier-phase dependence, control-chain distortion, readout imperfections). The Figure-8 waveform breaks the symmetry intrinsically through its 2:1 Fourier content; a third-order resonance ω+ω−2ω=0\omega+\omega-2\omega=0ω+ω−2ω=0 mediated by the anharmonic level ∣2⟩|2\rangle∣2⟩ predicts CF8∝A6N2C_{F8}\propto A^{6}N^{2}CF8​∝A6N2 at weak drive, confirmed by direct numerical propagation (fitted log-log slope 5.97±0.025.97\pm0.025.97±0.02; loop-number ratio 2434 vs. predicted 2500). On the successor device Cepheus-1-108Q, the predicted phase contrast lies below the binomial shot-noise floor at the accessible amplitudes, but the same dataset reveals a stronger signature of the underlying first-order average-Hamiltonian cancellation: the phase-averaged survival probability of the Figure-8 train remains pinned at ⟨P(∣0⟩)⟩ϕ≈0.985\langle P(|0\rangle)\rangle_{\phi}\approx0.985⟨P(∣0⟩)⟩ϕ​≈0.985 across all tested conditions, while the isenergetic rectangular reference sweeps the full range 0.24-0.93—an approximately 70-fold suppression of the Rabi-drive response. Together these results identify APC as a measurable control-landscape feature with distinct intrinsic and extrinsic components, and demonstrate a calibration-free, amplitude-insensitive idling property of Figure-8 pulse trains.
Closed-loop superconducting-qubit calibration has matured into DAG-orchestrated protocol chains, yet published frameworks treat the bath via a Markovian master equation or a phenomenological likelihood, absorbing bath structure into fit residuals instead of reporting it as a diagnostic. We integrate a QuTiP 5.x hierarchical-equations-of-motion (HEOM) solver driven by a Tier-1 1/f Burkard bath into a multi-protocol calibration DAG (Rabi -> {Ramsey || T1}) and benchmark it against sesolve and mesolve on a frozen platform in a pulse-level simulator (no hardware validation). The Ramsey channel carries the headline: the Markovian fit is censored by its exponential-family numerical ceiling, while HEOM recovers a physical revival envelope whose primary T2* separates from the Markovian reference by at least 13x at 95% independent-bootstrap confidence within the HEOM-feasible budget; the point-estimate ratio reaches >=28x on the 50-point primary-t1 grid and ~72x on the 30-point biexp-family tau_aw pivot at L=5. Rabi contrast falls 2.17% below mesolve on a noise-limited 30-point grid; the paired-bootstrap CI crosses zero, so this channel corroborates rather than independently establishes the non-Markovian signature. T1 decay shape matches across backends (beta=1.000), yet HEOM's initial occupation drops from 1.000 to 0.879 -- a bath-dressed contamination stable under a 16-point densification. The DAG adds 9.62 us average per-protocol scheduling overhead, no meaningful latency penalty at protocol granularity. HEOM-in-loop thereby changes what calibration reports: bath structure appears as a quantifiable residual rather than a hidden confound.
Every matter-based qubit platform—superconducting circuits, trapped ions, nitrogen- vacancy centres—exhibits decoherence in excess of what the standard two-level error model predicts from known noise sources. In superconducting qubits, two-level system (TLS) defect coupling accounts for 30–50% of the dephasing budget with debated mi- croscopic origin. In trapped ions, anomalous motional heating exceeds Johnson noise predictions by two to six orders of magnitude. In NV centres, residual pure dephasing persists after the nuclear spin bath and all identified sources are eliminated. These phenomena are platform-specific in detail but share a common structural feature: they all involve coupling of the qubit’s spin-projection state to spatial or motional degrees of freedom that the two-level model does not describe. Photonic qubits, by contrast, exhibit none of these anomalies. We propose that this asymmetry has a topological origin. A recent classifica- tion of electromagnetic field configurations into open (Class O, massless) and closed (Class C, massive) Poynting topologies [1] implies that massive particles carry internal structure—chirality, spatial mode geometry, Beltrami eigenstate classification—beyond their spin projection. Photons do not. The anomalous decoherence channels observed in every matter-based platform, and absent in every photonic platform, are the envi- ronmental coupling to this internal topological structure that the standard qubit model omits. This is not a small correction: the topological channels account for 5–50% of the observed decoherence, depending on the platform, and their platform-specific coupling fractions are consistent with known spin-to-spatial-mode coupling mechanisms at each scale.