Efficient calibration and control of programmable photonic integrated circuits are fundamental for scaling quantum and classical optical computing processors. While neural network-based models offer an architecture-agnostic solution, existing approaches suffer from limited learning and generalization capabilities due to the many-to-one mapping problem between sets of control signals and optical responses, and biased training datasets derived from uniform current sampling. In this work, we propose a universal calibration and control framework employing tandem neural networks combined with two novel data generation strategies: architecture-aware sampling based on Haar measure principles, and optimized sampling, a physics-agnostic approach utilizing differential evolution. We experimentally validate these methods on 3x3 and 4x4 coherent MZI meshes, demonstrating that our approach addresses the sampling bias inherent in previous works. When evaluated using random unitary matrices, our solution outperforms standard uniform sampling baselines by ~2 bits of precision. Furthermore, we experimentally extend the application of this framework to coherent detection, achieving precise control over both amplitude and phase, and validate its impact on photonic neural network tasks.
Efficient calibration and control of programmable photonic integrated circuits are fundamental for scaling quantum and classical optical computing processors. While neural network-based models offer an architecture-agnostic solution, existing approaches suffer from limited learning and generalization capabilities due to the many-to-one mapping problem between sets of control signals and optical responses, and biased training datasets derived from uniform current sampling. In this work, we propose a universal calibration and control framework employing tandem neural networks combined with two novel data generation strategies: architecture-aware sampling based on Haar measure principles, and optimized sampling, a physics-agnostic approach utilizing differential evolution. We experimentally validate these methods on 3x3 and 4x4 coherent MZI meshes, demonstrating that our approach addresses the sampling bias inherent in previous works. When evaluated using random unitary matrices, our solution outperforms standard uniform sampling baselines by ~2 bits of precision. Furthermore, we experimentally extend the application of this framework to coherent detection, achieving precise control over both amplitude and phase, and validate its impact on photonic neural network tasks.
Universal multiport interferometers based on Mach–Zehnder meshes are becoming a core building block for photonic neural-network accelerators and photonic quantum processors. This thesis presents the design, calibration, and control of such a programmable photonic integrated circuit (PIC). A fully automated software and hardware platform is developed to characterize and control the tunable elements of the PIC, enabling robust implementation of arbitrary linear optical transformations as well as target unitary operations. Using the calibrated device model, the control framework programs the PIC to realize specific unitaries and evaluates their performance through fidelity metrics computed from measured transfer matrices. By covering the complete stack from chip design to automated calibration and high-level unitary programming, this work establishes a scalable framework for design and operation of programmable photonic processors.
The growing scale and complexity of photonic quantum systems demand a transition from manual, heuristic control toward intelligent, autonomous operational frameworks. Quantum Artificial Intelligence (QAI) integrates machine learning, quantum-enhanced optimization, and real-time feedback to address key challenges in the calibration, stabilization, and scaling of photonic quantum technologies. Unlike conventional machine-learning-assisted quantum control and self-driving laboratory approaches that rely primarily on classical intelligence, the proposed QAI paradigm establishes a tighter integration between artificial intelligence and quantum hardware. Parameterized quantum circuits executed directly on the photonic processor participate in the agent’s decision-making process, while the same framework supports both real-time control tasks. QAI combines variational quantum neural networks, meta-reinforcement learning, multi-agent control architectures, and quantum optimization methods, such as the Quantum Approximate Optimization Algorithm, to enable adaptive calibration, error mitigation, and high-dimensional optimization in programmable photonic platforms. While this perspective focuses on real-time control and calibration, we also outline a conceptual pathway toward autonomous generative inverse design. To guide future development, we propose a three-stage roadmap that progresses from open-source hybrid quantum–classical control tools to fully autonomous laboratories capable of closed-loop self-optimization and experiment design. Experimental platforms include programmable silicon photonic circuits and lithium-niobate interferometers, where thermal drift, fabrication variability, and crosstalk require intelligent adaptation. Beyond quantum optics, the QAI framework has broader implications for quantum sensing, secure communications, and photonic computing. By introducing a unified QAI framework and a structured classification of Quantum Autonomy Levels (QALs), this work provides a roadmap toward autonomous quantum-optics laboratories and photonic quantum technologies.
The growing computational demands of classical neural networks have intensified the search for energy-efficient and powerful computational alternatives. Quantum neural networks (QNNs) implemented on integrated photonic platforms offer a compelling avenue, offering exceptional computational power enhancements, with inherent programmability and scalability of integrated architectures. A critical challenge, however, is implementing the fundamental non-unitary and nonlinear activation function of QNNs within a linear quantum photonic system. Existing strategies, such as the adding ancillary qubits and measurement-based feedback or forward are constrained by high qubit resource costs, overhead devices, and poor cascadability. Here, we propose a novel deep photonic QNN with an expanded computational Hilbert space via input replication and mode expansion, which enables the realization of effective non-unitary and nonlinear activation on a linear programmable quantum photonic chip. This approach eliminates the need for physical ancillary qubits, measurement-induced qubit consumption and the measurement device burden, thereby significantly reduce resource costs. The fabricated chip integrates four high-quality entanglement sources and a programmable high-dimensional interferometric network, enabling a two-hidden-layer QNN that exhibits dimension-enhanced expressivity over the existing QNN architectures. We demonstrate its capabilities across diverse tasks, including nonlinear classification, image generation, and quantum Gibbs state preparation. This work establishes a scalable and efficient architecture toward practical quantum deep learning systems capable of tackling problems beyond the reach of classical computation.
The rapid advancements in quantum computing (QC) and machine learning (ML) have led to the emergence of quantum machine learning (QML), which integrates the strengths of both fields. Among QML approaches, variational quantum circuits (VQCs), also known as quantum neural networks (QNNs), have shown promise both empirically and theoretically. However, their broader adoption is hindered by reliance on quantum hardware during inference. Hardware imperfections and limited access to quantum devices pose practical challenges. To address this, the Quantum-Train (QT) framework leverages the exponential scaling of quantum amplitudes to generate classical neural network parameters, enabling inference without quantum hardware and achieving significant parameter compression. Yet, designing effective quantum circuit architectures for such quantum-enhanced neural programmers remains non-trivial and often requires expertise in quantum information science. In this paper, we propose an automated solution using differentiable optimization. Our method jointly optimizes both conventional circuit parameters and architectural parameters in an end-to-end manner via automatic differentiation. We evaluate the proposed framework on classification, time-series prediction, and reinforcement learning tasks. Simulation results show that our method matches or outperforms manually designed QNN architectures. This work offers a scalable and automated pathway for designing QNNs that can generate classical neural network parameters across diverse applications.
Photonic integrated circuits offer a compact and stable platform for generating, manipulating, and detecting light. They are instrumental for classical and quantum applications. Imperfections stemming from fabrication constraints, tolerances and operation wavelength impose limitations on the accuracy and thus utility of current photonic integrated devices. Mitigating these imperfections typically necessitates a model of the underlying physical structure and the estimation of parameters that are challenging to access. Direct solutions are currently lacking for mesh configurations extending beyond trivial cases. We introduce a scalable and innovative method to characterize photonic chips through an iterative machine learning-assisted procedure. Our method is based on a clear-box approach that harnesses a fully modeled virtual replica of the photonic chip to characterize. The process is sample-efficient and can be carried out with a continuous-wave laser and powermeters. The model estimates individual passive phases, crosstalk, beamsplitter reflectivity values and relative input/output losses. Building upon the accurate characterization results, we mitigate imperfections to enable enhanced control over the device. We validate our characterization and imperfection mitigation methods on a 12-mode Clements-interferometer equipped with 126 phase shifters, achieving beyond state-of-the-art chip control with an average 99.77 % amplitude fidelity on 100 implemented Haar-random unitary matrices.
We introduce a distributed quantum-classical framework that synergizes photonic quantum neural networks (QNNs) with matrix product state (MPS) mapping to achieve parameter-efficient training of classical neural networks. By leveraging universal linear-optical decompositions ofM-mode interferometers and photon-counting measurement statistics, our architecture generates neural parameters through a hybrid quantum-classical workflow: photonic QNNs withM(M+1)/2trainable parameters produce high-dimensional probability distributions that are mapped to classical network weights via an MPS model with bond dimensionχ. Empirical validation on MNIST classification demonstrates that photonic QT achieves an accuracy of95.50%±0.84%using 3,292 parameters (χ=10), compared to96.89%±0.31%for classical baselines with6,690parameters. Moreover, a ten-fold compression ratio is achieved atχ=4, with a relative accuracy loss of less than 3 %. The framework outperforms classical compression techniques (weight sharing/pruning) by6−12%absolute accuracy while eliminating quantum hardware requirements during inference through classical deployment of compressed parameters. Simulations incorporating realistic photonic noise demonstrate the framework's robustness to near-term hardware imperfections. Ablation studies confirm quantum necessity - replacing photonic QNNs with random inputs collapses accuracy to chance level(10.0%±0.5%). Photonic quantum computing room-temperature operation, inherent scalability through spatial mode multiplexing, and HPC-integrated architecture establish a practical pathway for distributed quantum machine learning, combining the expressivity of photonic Hilbert spaces with the deployability of classical neural networks.