Scaling superconducting quantum processors is fundamentally limited by environmentally induced decoherence arising from nanoscale fabrication defects. Conventional heuristic microwave control protocols cannot dynamically adapt to device-specific material variations, while purely data-driven deep learning approaches often violate physical constraints. Here we introduce a compact quantum physics-informed neural network (QPINN) that learns device-aware microwave control pulses for superconducting qubits. Trained on empirical defect characteristics extracted from 162 InAs nanosheet Josephson junction measurements, the model embeds the Lindblad master equation directly into its loss function to enforce physical consistency. Open-system simulations demonstrate a 42% reduction in leakage to non-computational states. Hardware experiments executed on the 156-qubit IBM Quantum processor show an increase in Ramsey dephasing time from 83.3 ± 7.8 μs to 113.0 ± 1.1 μs while maintaining gate fidelity exceeding 99.99%.
Scaling superconducting quantum processors is fundamentally limited by environmentally induced decoherence arising from nanoscale fabrication defects. Conventional heuristic microwave control protocols cannot dynamically adapt to device-specific material variations, while purely data-driven deep learning approaches often violate physical constraints. Here we introduce a compact quantum physics-informed neural network (QPINN) that learns device-aware microwave control pulses for superconducting qubits. Trained on empirical defect characteristics extracted from 162 InAs nanosheet Josephson junction measurements, the model embeds the Lindblad master equation directly into its loss function to enforce physical consistency. Open-system simulations demonstrate a 42% reduction in leakage to non-computational states. Hardware experiments executed on the 156-qubit IBM Quantum processor show an increase in Ramsey dephasing time from 83.3 ± 7.8 μs to 113.0 ± 1.1 μs while maintaining gate fidelity exceeding 99.99%.
The potential of quantum computing is fundamentally constrained by the inherent susceptibility of qubits to noise and crosstalk, particularly during multi-qubit gate operations.
Existing strategies, such as hardware isolation and dynamical decoupling, face limitations in scalability, experimental feasibility, and robustness against complex noise sources.
In this manuscript, we propose a physics-guided neural control (PGNC) framework to generate robust control pulses for superconducting transmon qubit systems, specifically targeting crosstalk mitigation.
By combining a hardware aware parameterization with a Hamiltonian-informed objective that accounts for condition-dependent crosstalk distortions, PGNC steers the search toward smooth and physically realizable pulses while efficiently exploring high dimensional control landscapes.
Numerical simulations for the CZ gate demonstrate superior fidelity and pulse smoothness compared to a Krotov baseline under matched constraints.
Taken together, the results show consistent and practically meaningful improvements in both nominal and perturbed conditions, with pronounced gains in worst-case fidelity, supporting PGNC as a viable route to robust control on near-term transmon devices.
Quantum information processing holds the promise of a paradigm shift in computational power for specific problem classes. While contemporary quantum processors operate in the Noisy Intermediate-Scale Quantum (NISQ) regime and remain highly susceptible to errors, ongoing experimental efforts on these near-term devices are crucial for advancing the underlying technology. Among the various hardware implementations, superconducting quantum circuits stand out as one of the most robust architectures for scalable quantum technologies. Nevertheless, the primary bottleneck in these systems is the fragility of quantum information: interactions with the environment inevitably perturb the quantum states, a phenomenon known as decoherence. This rapid loss of stored information imposes a fundamental limit on the time available to reliably execute computations on these devices. Within this context, this thesis focuses on the comprehensive experimental characterization of a planar chip featuring isolated, fixed-frequency superconducting transmon qubits. The measurements were conducted in a cryogenic environment, relying on a dilution refrigerator to maintain the device at millikelvin temperatures. To probe the quantum system under these conditions, custom measurement scripts were developed to operate state-of-the-art microwave control electronics (Keysight Quantum Control System), establishing full control over pulse generation, signal acquisition, and digital demodulation. The experimental workflow systematically mapped the physical properties of the device through a progressive series of characterization steps. Initial spectroscopic techniques probed the coupled resonator-qubit system, identifying its fundamental transition frequencies and relevant energy scales. Building upon these frequency-domain measurements, coherent control was achieved by calibrating resonant microwave pulses to drive single-qubit rotations, enabling the reliable preparation of the qubit in excited or coherent superposition states. Finally, specific time-domain protocols were implemented to evaluate the operational limits of the device. By executing standard pulse sequences, the characteristic decoherence metrics of the system, namely the longitudinal energy relaxation time (T1) and the inhomogeneous dephasing time (T2*), were successfully extracted, providing a quantitative assessment of the qubit's overall performance and establishing a baseline for future implementations of more advanced quantum control protocols.
The classical Internet faces increasing limitations as data generation and device interconnectivity expand at unprecedented rates. In recent years, this has highlighted the need for communication frameworks that surpass classical limitations. The quantum Internet (QI), composed of multiple layers of quantum communication networks (QCNs), holds promise for remarkable capabilities in secure transmission, distributed quantum computing, and quantum-enhanced sensing. Accurate and scalable modeling of QCNs is essential for designing, analyzing, and optimizing QI protocols under realistic physical conditions. To model such open quantum systems under Markovian noise, the Lindblad master equation is widely used, but its modeling complexity grows exponentially with the system size. In this work, we introduce a learned approach that utilizes a physics-informed neural network (PINN) to model the dynamics of open quantum systems, which embeds the Lindblad formalism directly into the neural network's loss function. We demonstrate the effectiveness of our method by simulating the entanglement fidelity decay of a two-qubit Bell state under energy damping and phase decoherence noise, which is a relevant scenario for entanglement-based QI protocols. Preliminary numerical results show that the PINN-based solver aligns with conventional partial differential equation techniques while offering improved scalability. These findings suggest that PINNs could serve as an efficient approach for accelerating the analysis of future quantum networking solutions. Finally, we conclude by outlining key future directions for scaling the framework to support QCN-level optimization.
Fault-tolerant quantum computation requires scaling quantum processors to millions of physical qubits, yet purely superconducting architectures face significant hardware and control overhead. Hybrid multimode piezomechanical systems offer a hardware-efficient alternative by encoding quantum information in dense bosonic acoustic modes. However, stabilizing entanglement in such networks is challenging because continuous measurement and environmental dissipation generate strongly nonlinear driven-dissipative dynamics that are difficult to optimize analytically. Here we present an autonomous artificial-intelligence framework for the design and control of multimode superconducting optomechanical circuits. Using a model-free Recurrent Proximal Policy Optimization agent, the system learns a dynamic control protocol that exploits weak continuous measurement backaction to emulate a squeezed-vacuum environment, stabilizing steady-state remote entanglement with logarithmic negativity EN = 0.506 ± 0.027. In parallel, a Physics-Informed Neural Network incorporating the Lindblad master equation identifies control trajectories near a Liouvillian exceptional point, enabling efficient navigation of the non-Hermitian phase space. This physics-guided optimization reduces the circuit depth of standard quantum parity-check routines by 43.2% relative to conventional compilation. These results demonstrate that integrating machine learning with quantum control can autonomously discover robust strategies for stabilizing entanglement and reducing circuit complexity, providing a scalable pathway toward hardware-efficient quantum networks and distributed quantum information processing.
This work demonstrates real-time predictive classical-quantum feedback control on the IBM Kingston superconducting quantum processor, characterizing the hardware latency bottleneck of dynamic Hamiltonian shaping. Decoherence in quantum systems is traditionally viewed as a technical obstacle arising from environmental noise; this paper interprets decoherence as the coupling of a localized system to a dissipative, high-entropy environmental field. An ancilla-based modular sensor architecture is introduced to facilitate real-time error syndrome extraction without collapsing the target state. A software-defined controller, implemented via a quantized Variational Autoencoder-Feed-Forward Network (VAE-FFN) pipeline and optimized via empirical threshold calibration under strict 4-bit latency constraints, provides predictive Hamiltonian control to stabilize the system boundary. The integration of a 3D volumetric topological complex outlines a geometric architecture designed to mitigate spatial error cascades. Through temporal delay sweeps, empirical telemetry demonstrates that the volumetric lattice achieves localized containment of non-commuting phase defects, dynamically exploiting coherent Larmor precession to constrain spatial error propagation. Furthermore, rigorous statistical repetition mapping reveals a critical Noisy Intermediate-Scale Quantum (NISQ) latency bottleneck. The empirical data establishes that the temporal cost of classical-quantum feed-forward routing incurs a marginal macroscopic fidelity reduction (-1.01%). Crucially, this study reframes this differential not as a systemic regression, but as an explicit architectural trade-off: prioritizing rigorous variance suppression and operational stability over unmitigated peak fidelity. By establishing a highly predictable, predictively controlled baseline, this framework explores practical control strategies consistent with Approximate Quantum Error Correction (AQEC) approaches, bounding the operational consequences of the Eastin-Knill constraint within an addressable temporal control protocol.
The potential of quantum computing is constrained by the susceptibility of qubits to noise and crosstalk during multi-qubit gate operations. Existing strategies, such as hardware isolation and dynamical decoupling, face limitations in scalability, feasibility, and robustness against complex noise sources. Here we show a physics-guided neural control framework that generates robust control pulses for superconducting transmon qubit systems, specifically targeting crosstalk mitigation. By combining a hardware-aware parameterization with a Hamiltonian-informed objective that accounts for condition-dependent crosstalk distortions, the framework steers the search toward smooth and physically realizable pulses while exploring high-dimensional control landscapes. Numerical simulations for the controlled-Z gate demonstrate superior fidelity and pulse smoothness compared to the Krotov and GRAPE baselines under matched constraints. The results show consistent and practically meaningful improvements in both nominal and perturbed conditions, with pronounced gains in worst-case fidelity, supporting physics-guided neural control as a route to robust control on near-term transmon devices. Quantum computers are highly sensitive to unwanted crosstalk between qubits, especially during multi-qubit gate operations. Here, the authors show that physics-guided neural control can generate smooth, hardware-compatible pulses that make two-qubit gates more robust against crosstalk.