A quantum network utilizes quantum entanglement and quantum error correction to effectively transmit quantum information. The development of such networks is a crucial step toward achieving secure and efficient quantum communication. This paper introduces a Bell inequality to verify nonlocality in continuous-variable (CV) multipartite chained entanglement-swapping networks, and proposes a CV-quantum key distribution (QKD) protocol for such networks with Bell-test gating, in which key generation proceeds only when the Bell test is passed. We analyze the achievable key rates under collective and coherent attacks. Numerical simulations show a clear scalability advantage, with the maximum secure distance increasing with the number of nodes. These results support the proposed CV-QKD protocol as a practical route to long-distance secure communication in CV quantum networks.

Weak measurement (WM) offers the advantage of amplifying small signals at the cost of postselecting a small portion of the probes. This so-called weak-value amplification effect makes it compare favorably with conventional techniques (without postselection) to overcome various technical noises. However, certain types of technical noise, such as jitter and pixelation, present insurmountable limitations for both WM and conventional techniques. In this work, we propose an advanced variant of WM that incorporates time-momentum correlation (TMC) into biased weak measurement (BWM). By employing the Fisher information metric, we theoretically show that TMC-BWM can overcome jitter and pixelation, and meanwhile extract significantly higher Fisher information. This dual advantage was experimentally validated in a magnetic-sensing application when applying severe jitter and pixelation, and it remains robust under natural conditions, achieving a 24.8-dB improvement in precision over the standard WM scheme. Moreover, merely classical resources are required to achieve this dual metrological advantage.

Lowering the resource overhead needed to achieve fault-tolerant quantum computation is crucial to building scalable quantum computers. We show that adapting conventional maximum likelihood (ML) decoders to a small subset of efficiently learnable physical error characteristics can significantly improve the logical performance of a quantum error-correcting code. Specifically, we leverage error information obtained from efficient characterization methods based on cycle error reconstruction (CER), which yields Pauli error rates on the $n$ qubits of an error-correcting code. Although the total number of Pauli error rates needed to describe a general noise process is exponentially large in $n$, we show that only a few of the largest few Pauli error rates are needed and that a heuristic technique can complete the Pauli error distribution for ML decoding from this restricted dataset. Using these techniques, we demonstrate significant performance improvements for decoding quantum codes under a variety of physically relevant error models. For instance, with CER data that constitute merely 1% of the Pauli error rates in the system, we achieve a 10 times gain in performance compared with the case where decoding is based solely on the fidelity of the underlying noise process. Our conclusions underscore the promise of recent error characterization methods for improving quantum error correction and lowering overheads.

Artificial atoms nonlocally coupled to waveguides—the so-called giant atoms—offer opportunities for the control of light and matter. In this work, we show how to use an array of nonlocally coupled transmon “molecules” to engineer a passive photonic controlled gate for waveguide photons. In particular, we show that a conditional elastic phase shift between counterpropagating photons arises from the interplay between direction-dependent couplings, engineered through an interplay of nonlocal interactions and molecular binding strength; and the nonlinearity of the transmon array. We analyze the conditions under which a maximal $\pi$-phase shift—and hence a CZ gate—is obtained, and characterize the gate fidelity as a function of key experimental parameters, including finite transmon nonlinearities, emitter spectral inhomogeneities, and limited cooperativity. Our work opens the use of giant atoms as key elements of microwave photonic quantum computing devices.

Accurate ground-state calculations on noisy quantum computers are significantly constrained by restricted Ansatz expressivity and unavoidable hardware errors. We introduce a hybrid quantum-classical framework that simultaneously addresses these challenges. Our method systematically purifies noisy two-electron reduced density matrices from quantum devices by enforcing $N$-representability conditions through efficient semidefinite programming, guided by a norm-based distance constraint to the experimental data. To implement this constraint, we develop a hardware-efficient calibration protocol based on Clifford circuits. We demonstrate near full configuration-interaction accuracy for ground-state energies of ${\mathrm{H}}_2$, $\mathrm{Li}\mathrm{H}$, and ${\mathrm{H}}_4$, and compute precise scattering intensities for ${\mathrm{C}}_6{\mathrm{H}}_8$ on noisy hardware. This approach surpasses conventional methods by simultaneously overcoming both Ansatz limitations and hardware noise, providing a framework for quantum advantage and offering an alternative for reliable simulations of complex molecular systems on noisy devices.

Quantum machine learning (QML) has emerged as a promising framework for exploring how quantum dynamics may enhance data processing tasks. Here we investigate quantum extreme-learning machines (QELMs), a quantum analog of classical extreme-learning machines in which training is restricted to the output layer. Our architecture combines dimensionality reduction (via principal component analysis or autoencoders), quantum state encoding, evolution under an XX Hamiltonian, and projective measurement to produce features for a classical single-layer classifier. By analyzing the classification accuracy as a function of evolution time, we observe a sharp transition between low- and high-accuracy regimes, followed by saturation. Remarkably, the saturated performance is comparable to that obtained using Haar-random unitaries that generate maximally complex dynamics, even though the XX model is integrable and local. Our results indicate that this increase in performance correlates with the onset of entanglement, which improves the embedding of classical data in Hilbert space and leads to more separable clusters in measurement probability space. Thus, moderate entanglement can contribute positively to the structure of the data representation, improving learnability without necessarily implying quantum computational advantage. For the image-classification tasks studied here, namely the Modified National Institute of Standards and Technology (MNIST), Fashion-MNIST, and CIFAR-10 datasets, the relevant evolution time is consistent with information exchange over short distances and, within the explored system sizes, does not show evidence of scaling with the full system size. This suggests that QELM performance in this regime relies only on limited entanglement and remains compatible with efficient classical simulation. Our results clarify how local quantum dynamics and moderate quantum correlations are already sufficient to generate useful feature representations for learning.

State transitions during qubit measurements are extremely detrimental to quantum tasks that rely on repeated measurements, such as quantum error correction. These state transitions can occur when excessive measurement power leads to qubit excitations outside its computational space. Alternatively, the qubit state can decay rapidly when the measurement protocol inadvertently couples the qubit to lossy modes such as two-level systems (TLSs). We experimentally verify the impact of these TLSs in qubit readout by measuring the transition errors at different qubit flux bias. Because such state transitions during measurements are often localized in frequency space, we demonstrate the ability to avoid them during a fluxonium readout by exploiting the qubit’s flux tunability. By synchronizing the flux bias with the readout photon dynamics, we reduce the readout error by 48% (49%) in 1-$\text{μ}\mathrm{s}$ (0.5-$\text{μ}\mathrm{s}$) integration time. Our work advances the understanding of state transitions in superconducting circuit measurements and demonstrates the potential of fluxonium qubits to achieve fast high-fidelity readout.

Stochastic methods offer an effective way to suppress coherent errors in quantum simulation. In particular, the randomized compilation protocol may reduce circuit depth by randomly sampling Hamiltonian terms rather than following the deterministic Trotter-Suzuki sequence. However, its fixed sampling distribution does not adapt to the dynamics of the system, limiting its accuracy. In this work, we propose a fluctuation-guided adaptive algorithm that adaptively updates sampling probabilities based on fluctuations of Hamiltonian terms to achieve higher simulation fidelity. Remarkably, the protocol renders an intuitive physical understanding: Hamiltonian terms with greater sensitivity to the state evolution should be prioritized during sampling. The overhead of measuring fluctuations necessary for updating the sampling probability is affordable, and can be further largely reduced by classical shadows. We demonstrate the effectiveness of the method with numeral simulations across discrete-variable, continuous-variable, and hybrid-variable systems.

Measurement-device-independent quantum key distribution (MDI-QKD) eliminates all detector-side vulnerabilities. Compared to weak coherent pulses, heralded single-photon sources (HSPSs) signal the presence of photons, thereby extending the achievable distance of MDI-QKD and enabling advanced protocols such as passive decoy-state implementations and quantum-memory-assisted configurations. In practice, however, spectral correlations between photon pairs in HSPS degrade Hong-Ou-Mandel (HOM) interference visibility. Although narrowband filtering can suppress this effect, it relies on dedicated hardware and lacks flexibility. Here, we propose an MDI-QKD scheme based on a time-domain postprocessing filtering technique that requires no additional hardware, significantly improving both the key rate and the maximum transmission distance. By optimizing the filter width, this method flexibly adapts to variations in the source or channel. Furthermore, it can be readily extended to other protocols relying on HOM interference, including quantum teleportation and entanglement swapping.

We demonstrate a practical countermeasure against a well-known class of attacks on quantum key distribution (QKD) systems that exploit detection-efficiency mismatch, where the receiver’s detectors do not exhibit identical responses to incoming photons across all degrees of freedom. This class of quantum hacking strategies is broad and significantly includes the time-shift attack, which targets an arrival-time-dependent side-channel at the receiver. The four-state countermeasure, previously only proven to be secure in theory, is implemented here on a gigahertz-clocked prototype QKD system and evaluated for its security and performance. We show that its presence enables almost complete recovery of the system’s ideal secret key rate. Our results provide strong justification for adopting this countermeasure as a standard component in future scalable and practical QKD systems.

Block encoding is a key ingredient in the recently developed quantum singular value transformation (QSVT) framework, which provides a unifying description for many quantum algorithms. Although it was initially introduced to simplify and optimize resource utilization in various problems—such as searching, amplitude estimation, and Hamiltonian simulation—it is reasonable to expect that the capabilities of QSVT extend beyond these applications and offer untapped potential for designing new quantum algorithms. In this paper, we affirm this perspective by leveraging block encoding to substantially enhance two previously proposed quantum algorithms: largest eigenvalue estimation and quantum gradient descent. Unlike previous works that rely on sophisticated approaches, our findings demonstrate that even using just elementary operations within the unitary block-encoding framework can eliminate major scaling factors, present in their original counterparts. This results in more efficient quantum algorithms, which are capable of tackling target computational problems with better efficiency. Furthermore, we illustrate how our proposed method can be extended to other contexts, including matrix inversion and multiple eigenvalue estimation, and discuss its potential for a variety of applications toward the stability analysis of physical systems, mechanical vibration, and transport phenomena, as well as molecular geometry optimization.

Modular quantum architectures have emerged as a promising approach for scaling quantum computing systems by connecting multiple quantum processing units. However, this approach introduces significant challenges due to costly intercore operations between chips and quantum state transfers, which contribute to noise and quantum decoherence. This paper presents QARMA, a qubit-mapping approach using attention-based deep reinforcement learning for modular quantum architectures, along with its extension, QARMA-R, which incorporates dynamic qubit reuse. Our approach combines an attention-based mechanism with graph neural networks to learn optimal qubit allocation, routing, and reuse strategies that minimize intercore communication. Our experimental results show significant improvements over state-of-the-art approaches. Compared to highly optimized Qiskit with modular architecture configuration, QARMA-R reduces intercore communication by up to 100% (86% on average), while QARMA maintains 15%–40% improvement for larger circuits without reuse. Against traditional modular qubit mapping, our approach achieves a 97%–100% reduction in intercore operation. The proposed methods advance quantum circuit compilation and enable the execution of larger quantum algorithms on resource-constrained modular quantum systems, thereby contributing to the growing body of research on scalable quantum computing architectures.

Quantum random-number generators (QRNGs) produce true random numbers based on the inherent randomness of quantum theory, rendering them a foundational segment of quantum cryptography. Distinguished from trusted-device QRNGs, whose security depends on characterized devices, semi-device-independent (semi-DI) QRNGs permit partial devices to be defective or even maliciously manipulated, achieving a good trade-off between generation rate and security. In this paper, we propose a semi-DI QRNG that resists general attacks while accounting for finite-size effects. The protocol requires no rigorous characterization of the source and measurement devices other than limiting the energy of the emitted states, significantly reducing the demand on practical QRNG systems. Leveraging the tight Kato’s inequality for correlated variables, we show that our protocol generates more randomness than it consumes. Furthermore, we demonstrate the scheme using a continuous-variable system with ternary inputs of states. Heterodyne detection is employed to enable phase compensation through data postprocessing, alleviating the stringent requirement on system stability. The system operates at 100 MHz, achieving a net random-number generation rate of 1.165 Mbps at $5.3\times {10}^9$ rounds. Our work offers a promising approach to achieving both robust security and a high generation rate with a simple experimental setup.

For the ever-growing field of quantum information processing, large-scale, efficient multiport interferometers serving as photonic processors are required. In this context, the suitability of quantum walks as the interferometric base for universal computation has been theoretically proven. In this work, we bridge the gap between theoretical proposals and state-of-the-art experimental capabilities by providing the recipe for the implementation of a universal photonic processor in discrete-time quantum walks. Specifically, we present the protocol for translating arbitrary linear transformations into the coin and step operator of a quantum walk and map these to the experimental parameters of the established time-multiplexed platform [A. Schreiber et al., Phys. Rev. Lett. 104, 050502 (2010)]. We show that our interface is highly scalable and resource efficient due to the hybrid encoding consisting of multiple degrees of freedom. Finally, we prove that our system is highly resilient against experimental imperfections and show that it compares favorably against existing architectures.

Fast, high-fidelity, and low-backaction readout plays a crucial role in the advancement of quantum error correction (QEC). Here we demonstrate high-performance multiplexed readout of superconducting qubits using a tunable broadband Purcell filter, effectively resolving the fundamental trade-off between measurement speed and photon noise–induced dephasing. By dynamically tuning the filter parameters, we suppress photon noise–induced dephasing by a factor of 7 in idle status, while enabling rapid, high-fidelity readout during measurement. Using 100 ns readout pulses, we achieve low crosstalk simultaneous readout of three qubits with a relaxation-limited average fidelity of 99.5%. By using a multilevel readout technique, we overcome the relaxation constraint and achieve a fidelity of 99.9% with a 50 ns readout pulse. The system also exhibits high-performance quantum-nondemolition characteristics, maintaining 99.4% fidelity over repeated measurements with a qubit state leakage rate below 0.1%. Building on the tunable filter, we further propose a scalable readout scheme for surface code implementation with enhanced multiplexing capability, providing a promising pathway toward fast and scalable QEC.

We investigate how feature-space dimensionality reshapes relaxation and correlation structure in a transverse-field Ising model with feature-conditioned, ferromagnetically biased bond disorder, analyzed via its anisotropic (two plus one)-dimensional classical counterpart. Using Suzuki–Trotter simulations of the mapped classical system, we study imaginary-time autocorrelators, coarse equal-time spatial summaries, and eigenvalue spectra of equal-time correlation matrices. Each spin is assigned a $d$-component feature vector, and a similarity scale $\hat{\sigma }$ controls how feature differences bias the distribution of bond signs. Increasing $d$ homogenizes pairwise feature distances and, together with $\hat{\sigma }$, tunes bond-sign heterogeneity within this positively biased ensemble. We find a systematic dynamical crossover from short-$\tau$ transients at low $d$ to multistep relaxation with long-lived plateaus at higher $d$, accompanied by a compression of equal-time correlation weight into fewer dominant spatial modes. Feature-conditioned disorder thus provides a controlled framework for studying how correlated disorder reshapes relaxation spectra in disordered Ising networks, with potential relevance to programmable Ising hardware.

Quantum reservoir computing is a machine-learning framework that offers ease of training compared to other quantum neural networks, as it does not rely on gradient-based optimization. Learning is performed in a single step on the output features measured from the quantum system. Various implementations of quantum reservoir computing have been explored in simulations, with different measured features. Although simulations have shown that quantum reservoirs present advantages in performance compared to classical reservoirs, experimental implementations have remained scarce. This is due to the challenge of obtaining a large number of output features that are nonlinear transformations of the input data. In this work, we show that even with a circuit quantum electrodynamics system as simple as a single transmon coupled to a readout resonator, we can implement a proof-of-concept realization of quantum reservoir computing. We obtain a large number of nonlinear features from a single physical system by encoding the input data in the amplitude of a coherent drive and measuring the cavity state in the Fock basis. We demonstrate classification of two classical tasks with significantly smaller hardware resources and fewer measured features compared to classical neural networks. Our experimental results are supported by numerical simulations that show additional Kerr nonlinearity is beneficial to reservoir performance. Our work demonstrates a hardware-efficient quantum neural-network implementation that can be further scaled up and generalized to other quantum machine-learning models.

Efficient and precise quantum state detection is essential for quantum information processing, serving as the backbone for high-fidelity initialization, real-time error correction, and scalable quantum operations. In particular, fluorescence-based detection, which is widely used in both neutral-atom and trapped-ion systems, must retain high fidelity under varying imaging conditions. Here, we present an adaptive convolutional neural network (CNN) that incorporates exposure time as a contextual input, enabling robust generalization across different exposure settings without retraining. To meet the stringent requirements of real-time quantum applications, we accelerate the model using a field-programmable gate array (FPGA), achieving an inference latency of $9.85\text{μ}\mathrm{s}$ at a 500 MHz clock frequency. Compared with state-of-the-art CPU and graphics processing unit (GPU) implementations, our hardware-accelerated design achieves an inference latency reduction of about $30\times$. These results highlight the potential for real-time context-aware quantum state detection, with a scalable model design that can incorporate additional experimental parameters. The proposed approach is well suited for integration in heterogeneous control architectures, supporting low-latency, resource-efficient, hybrid quantum-classical workflows.

With rapid progress being made in the development of platforms for quantum computation, there has been considerable interest in whether present-day and near-term devices can be used to solve problems of relevance. A commonly cited application area is the domain of quantum chemistry. While most experimental demonstrations of quantum chemical calculations on quantum devices have focused on the ground-state electronic energy of the system, other properties of the ground state, such as the electric dipole moment, are also of interest. Here, we employ the quantum-computed-moments method, based on the Lanczos cluster expansion, to estimate the dipole moment of the water molecule on an IBM Quantum superconducting quantum device. The noise-mitigated results agree with full-configuration-interaction calculations to within $0.03\pm 0.007$ D ($2\mathrm{\%}\pm 0.5\mathrm{\%}$), compared with direct expectation-value determination (i.e., variational quantum eigensolver, VQE) with errors of the order of 0.07 D (5%), even when the VQE calculation is performed without noise. This demonstrates that moments-based energy-estimation techniques can be adapted to noise-robust evaluation of nonenergetic ground-state properties of chemical systems. We emphasize that the dipole moment is chosen as an example of such a property, but the method applies to any property determined by one- or two-body interactions.