Abstract Integrated photonic circuits play a crucial role in implementing quantum information processing in the noisy intermediate‐scale quantum (NISQ) era. Variational learning is a promising avenue that leverages classical optimization techniques to enhance quantum advantages on NISQ devices. However, most variational algorithms are circuit‐model‐based and encounter challenges when implemented on integrated photonic circuits, because they involve explicit decomposition of large quantum circuits into sequences of basic entangled gates, leading to an exponential decay of success probability due to the non‐deterministic nature of photonic entangling gates. Here, a variational learning approach is presented for designing quantum photonic circuits, which directly incorporates post‐selection and elementary photonic components into the training process. The complicated circuit is treated as a single nonlinear logical operator and a unified design is discovered for it through variational learning. Engineering an integrated photonic chip with automated control achieved by genetic algorithm, the internal parameters of the chip are adjusted and optimized in real‐time for task‐specific cost functions. A simple case of designing photonic circuits for a single ancilla CNOT gate with improved success rate is utilized to illustrate how the proposed approach works, and then the approach is applied to the first demonstration of quantum stochastic simulation using integrated photonics.
Abstract Integrated photonic circuits play a crucial role in implementing quantum information processing in the noisy intermediate‐scale quantum (NISQ) era. Variational learning is a promising avenue that leverages classical optimization techniques to enhance quantum advantages on NISQ devices. However, most variational algorithms are circuit‐model‐based and encounter challenges when implemented on integrated photonic circuits, because they involve explicit decomposition of large quantum circuits into sequences of basic entangled gates, leading to an exponential decay of success probability due to the non‐deterministic nature of photonic entangling gates. Here, a variational learning approach is presented for designing quantum photonic circuits, which directly incorporates post‐selection and elementary photonic components into the training process. The complicated circuit is treated as a single nonlinear logical operator and a unified design is discovered for it through variational learning. Engineering an integrated photonic chip with automated control achieved by genetic algorithm, the internal parameters of the chip are adjusted and optimized in real‐time for task‐specific cost functions. A simple case of designing photonic circuits for a single ancilla CNOT gate with improved success rate is utilized to illustrate how the proposed approach works, and then the approach is applied to the first demonstration of quantum stochastic simulation using integrated photonics.
Integrated photonic circuits play a crucial role in implementing quantum information processing in the noisy intermediate-scale quantum (NISQ) era. Variational learning is a promising avenue that leverages classical optimization techniques to enhance quantum advantages on NISQ devices. However, most variational algorithms are circuit-model-based and encounter challenges when implemented on integrated photonic circuits, because they involve explicit decomposition of large quantum circuits into sequences of basic entangled gates, leading to an exponential decay of success probability due to the non-deterministic nature of photonic entangling gates. Here, we present a variational learning approach for designing quantum photonic circuits, which directly incorporates post-selection and elementary photonic elements into the training process. The complicated circuit is treated as a single nonlinear logical operator, and a unified design is discovered for it through variational learning. Engineering an integrated photonic chip with automated control, we adjust and optimize the internal parameters of the chip in real time for task-specific cost functions. We utilize a simple case of designing photonic circuits for a single ancilla CNOT gate with improved success rate to illustrate how our proposed approach works, and then apply the approach in the first demonstration of quantum stochastic simulation using integrated photonics.
Quantum embedding is a key technique in quantum machine learning, mapping classical data into a high-dimensional Hilbert space through parameterized quantum circuits to enhance data separability. While effective on general quantum devices, its implementation on photonic platforms faces severe challenges, as entangling gates rely on post-selection and lead to exponentially decreasing success probabilities when cascaded. To address this, we propose a quantum embedding learning approach tailored for photonic quantum circuits, which incorporates multi-photon post-selection into a variational ansatz, enabling the efficient generation of entanglement while maintaining circuit scalability on NISQ devices. By training the transfer matrices, our method achieves the desired photon distribution probabilities and input-output correspondence. We demonstrate its versatility through three tasks: Bell-state projection, quantum support vector machines (qSVM), and quantum clustering of K-means. The experimental results indicate enhanced performance in state discrimination, classification, and clustering, showing that the method provides a practical route to implement quantum embedding in integrated photonic circuits.
In the era of noisy intermediate-scale quantum computers, variational quantum algorithms are promising approaches for solving optimization tasks by training parameterized quantum circuits with the aid of classical routines informed by quantum measurements. In this context, photonic platforms based on reconfigurable integrated optics are an ideal candidate for implementing these algorithms. Among various techniques to train variational circuits, the parameter shift rule enables the exact calculation of cost function derivatives efficiently, facilitating gradient descent-based optimization. In this paper, we derive a formulation of the parameter shift rule for computing derivatives and integrals tailored to reconfigurable optical linear circuits and based on the Boson Sampling paradigm. This allows us to naturally embed common types of experimental noise, such as partial distinguishability and mixedness of the states, thus obtaining a resilient approach. Finally, we employ the developed approach to experimentally test variational algorithms with single-photon states processed in a reconfigurable six-mode universal integrated interferometer. Specifically, we apply the photonic parameter shift rules to the variational implementation, on a photonic platform, of both an eigensolver and a Universal- gate.
Quantum computing is among the most promising emerging techniques to solve problems that are computationally intractable on classical hardware. A large body of existing works focus on using variational quantum algorithms on the gate level for machine learning tasks, such as the variational quantum circuit (VQC). However, VQC has limited flexibility and expressibility due to limited number of parameters, e.g. only one parameter can be trained in one rotation gate. On the other hand, we observe that quantum pulses are lower than quantum gates in the stack of quantum computing and offers more control parameters. Inspired by the promising performance of VQC, in this paper we propose variational quantum pulses (VQP), a novel paradigm to directly train quantum pulses for learning tasks. The proposed method manipulates variational quantum pulses by pulling and pushing the amplitudes of pulses in an optimization framework. Similar to variational quantum algorithms, our framework to train pulses maintains the robustness to noise on Noisy Intermediate-Scale Quantum (NISQ) computers. In an example task of binary classification, VQP learning achieves up to 11% and 9% higher accuracy compared with VQC learning on the qiskit pulse simulators (with system model from real machine) and ibmq-jarkata, respectively, demonstrating its effectiveness and feasibility. Stability for VQP to obtain reliable results has also been verified in the presence of noise.
The use of Variational Quantum Circuits (VQCs), which are an important type of model in Quantum Machine Learning (QML) that will be used with Basic Quantum Computers during their Noisy intermediate-scale quantum (NISQ) era, is very dependent on the properties of the circuits themselves (architectures) and how they are trained (training settings). In fact, this model is so dependent on the knowledge of the human designer that creating these models takes significant time and energy to develop properly through qualitatively high amounts of quantum knowledge. In this work, we propose a new method for automatically building and training quantum machine learning models using an AutoML-Assisted approach to optimize VQCs. The overall goal of the AutoML-Assisted approach is to treat the design of quantum circuits as a learning problem and utilize Neural Architecture Search ( NAS ) to search end to end on the architecture, variational parameters, and hyperparameters automatically. Our experimental results on classification tasks show that the automatically learned VQCs not only give better accuracy but also give faster convergence and generally are more robust to quantum noise than warehouse VQCs developed by humans. This work makes a direct contribution to increasing the automation and scalability by which quantum machine learning can be run on NISQ computers.
Multi‐objective optimization in quantum systems poses formidable challenges stemming from the high dimensionality, complex constraints, and conflicting performance metrics inherent to quantum control tasks and circuit synthesis. This study presents a comprehensive exploration of genetic AI algorithms—evolutionary strategies inspired by natural selection—tailored specifically for multi‐objective optimization in quantum computing contexts. We propose a hybrid framework that synergistically combines the elitist, Pareto‐based selection mechanics of NSGA-II with quantum‐inspired crossover and mutation operators designed to preserve entanglement structures and exploit hardware‐specific gate sets. In this framework, candidate solutions encoding circuit topologies or control‐pulse parameters undergo iterative evolution, guided by fitness vectors that balance gate fidelity, circuit depth, and resource overhead. To benchmark performance, we apply the algorithm to canonical quantum computing problems: Fourier transform circuits, QAOA ansätze for MaxCut instances, and random Clifford+T circuits. Through extensive simulation on noise-modelled quantum devices, we evaluate Pareto front coverage, convergence speed, diversity maintenance, and robustness under decoherence. Our results demonstrate that the proposed genetic AI algorithm outperforms classical single‐objective genetic methods and established multi‐objective evolutionary algorithms (e.g., SPEA2) by achieving broader Pareto fronts and faster convergence rates, while maintaining solution diversity. Furthermore, noise-resilience experiments show graceful degradation, indicating practical applicability on near-term quantum hardware. These findings substantiate the potential of evolutionary computation paradigms—augmented with quantum‐inspired operators—to address the intricate trade-off surfaces that define NISQ-era optimization problems, paving the way toward automated quantum circuit design and control parameter tuning for real-world quantum computing applications.