Hybrid Quantum Neural Networks (HQNNs) combine classical learning with parameterized quantum circuits, but their practical performance is often limited by (i) the noise of Noisy Intermediate-Scale Quantum (NISQ) devices and (ii) the large, discrete design space of quantum circuit architectures. Moreover, HQNNs are commonly trained using a fixed circuit and a single backend, even though deployment frequently targets heterogeneous backends where compilation and execution characteristics may differ. To address these challenges, we propose GAT-QNN, a genetic algorithm (GA)-based framework that trains a macroCircuit (search space) by iteratively sampling microCircuits (subcircuits), training them, and reintegrating their learned parameters into the macroCircuit. After training, we run an independent GA-driven inference stage that evaluates candidate microCircuits using the trained macroCircuit weights and selects top-performing architectures for deployment. This two-stage approach enables backend-aware microCircuit selection without retraining each candidate architecture and can also reduce computational resources (gate count) by deploying smaller microCircuits derived from the macroCircuit. We validate the approach on MNIST classification (four classes) and report consistent 22-23% test accuracy gains for GA-driven inference across multiple backends.
Journal reference
2026 International Joint Conference on Neural Networks (IJCNN), Maastricht, Netherlands, 2026, pp. 1-8
Hybrid Quantum Neural Networks (HQNNs) combine classical learning with parameterized quantum circuits, but their practical performance is often limited by (i) the noise of Noisy Intermediate-Scale Quantum (NISQ) devices and (ii) the large, discrete design space of quantum circuit architectures. Moreover, HQNNs are commonly trained using a fixed circuit and a single backend, even though deployment frequently targets heterogeneous backends where compilation and execution characteristics may differ. To address these challenges, we propose GAT-QNN, a genetic algorithm (GA)-based framework that trains a macroCircuit (search space) by iteratively sampling microCircuits (subcircuits), training them, and reintegrating their learned parameters into the macroCircuit. After training, we run an independent GA-driven inference stage that evaluates candidate microCircuits using the trained macroCircuit weights and selects top-performing architectures for deployment. This two-stage approach enables backend-aware microCircuit selection without retraining each candidate architecture and can also reduce computational resources (gate count) by deploying smaller microCircuits derived from the macroCircuit. We validate the approach on MNIST classification (four classes) and report consistent 22-23% test accuracy gains for GA-driven inference across multiple backends.
We introduce Genetic Transformer Assisted Quantum Neural Networks (GTQNNs), a hybrid learning framework that combines a transformer encoder with a shallow variational quantum circuit and automatically fine tunes the circuit via the NSGA-II multi objective genetic algorithm. The transformer reduces high-dimensional classical data to a compact, qubit sized representation, while NSGA-II searches for Pareto optimal circuits that (i) maximize classification accuracy and (ii) minimize primitive gate count an essential constraint for noisy intermediate-scale quantum (NISQ) hardware. Experiments on four benchmarks (Iris, Breast Cancer, MNIST, and Heart Disease) show that GTQNNs match or exceed state of the art quantum models while requiring much fewer gates for most cases. A hybrid Fisher information analysis further reveals that the trained networks operate far from barren plateaus; the leading curvature directions increasingly align with the quantum subspace as the qubit budget grows, confirming that the transformer front end has effectively condensed the data. Together, these results demonstrate that GTQNNs deliver competitive performance with a quantum resource budget well suited to present-day NISQ devices.
Hybrid quantum-classical neural networks (HQNNs) are emerging as a practical approach for quantum machine learning in the noisy intermediate-scale quantum (NISQ) era, as they combine classical learning components with parameterized quantum circuits in an end-to-end trainable framework. However, their performance and efficiency depend strongly on architectural choices such as data encoding, circuit structure, measurement design, and the coupling between classical and quantum modules. This makes manual design increasingly difficult, especially when hardware limitations and resource constraints must also be taken into account. In this paper, we study the foundations of HQNNs and neural architecture search (NAS), discuss how NAS extends to quantum and hybrid settings, and demonstrate FLOPs-aware search (where FLOPs serve as a proxy for computational complexity), as an important hardware-aware direction for building HQNNs that are not only accurate but also computationally efficient and practically deployable.
Post-variational quantum neural networks (PV-QNNs) shift learning from parameterized quantum circuits to a classical head, but their performance hinges on selecting effective fixed ansätze. Manual templates or random circuit picks are brittle and scale poorly. We propose an automated, genetic-algorithm (GA) framework for PV-QNN ansatz ensemble selection. The GA searches a discrete circuit grammar (single-/two-qubit gates under topology and depth constraints) to assemble small ensembles whose quantum features are extracted via classical shadows and consumed by a shallow classical classifier. On Fashion-MNIST and MNIST (2 vs 6), GA-selected ensembles consistently outperform random and heuristic baselines while using about 40% fewer circuits; our efficiency metric, accuracy per circuit, improves by 74%. We also observe stable cross-dataset transfer, indicating that the discovered circuits capture task-relevant structure beyond a single training distribution. These results demonstrate that evolutionary search is a practical route to high-utility, low-footprint PV-QNNs, improving accuracy and sample efficiency without any quantum parameter training, and strengthening PV-QNN viability on near-term hardware.
Photonic quantum computing is a promising platform for scalable quantum machine learning, but designing effective hybrid architectures remains challenging under hardware and optimization constraints. Existing approaches rely on manually tuned architectures that fail to account for the collaboration between classical preprocessing, phase encoding, and photonic circuit structure, limiting both accuracy and hardware compatibility. In this paper, we propose a neural architecture search framework for hybrid photonic quantum-classical models that combines genetic algorithm-based search with learnable quantum phase encoding to systematically explore the joint design space of classical and quantum components. Our framework encodes 19 hyperparameters across six gene groups and evolves a population of hybrid architectures using group-based crossover, per-gene mutation, and elitism, evaluating each candidate on a short training budget before full retraining of the best found design. We evaluate our framework on two image classification benchmarks, Digits and MNIST, achieving final validation accuracies of 99.44% and 98.78%, respectively, with first-principles execution time estimates on the Quandela Ascella photonic QPU projecting single-image inference at ~67 ms (Digits) and ~149 ms (MNIST). Our quantum contribution analysis further shows that the photonic layer extracts non-redundant features orthogonal to the classical pathway, providing a measurable accuracy advantage over classical-only baselines. Our results demonstrate that automated architecture search is both practical and impactful for hybrid photonic systems, opening the way for systematic design space exploration of quantum AI on photonic devices.
Hybrid Quantum Neural Networks (HQNNs), which combine parameterized quantum circuits with classical neural layers, are emerging as promising models in the noisy intermediate-scale quantum (NISQ) era. While quantum circuits are not naturally measured in floating point operations (FLOPs), most HQNNs (in NISQ era) are still trained on classical simulators where FLOPs directly dictate runtime and scalability. Hence, FLOPs represent a practical and viable metric to measure the computational complexity of HQNNs. In this work, we introduce FAQNAS, a FLOPs-aware neural architecture search (NAS) framework that formulates HQNN design as a multi-objective optimization problem balancing accuracy and FLOPs. Unlike traditional approaches, FAQNAS explicitly incorporates FLOPs into the optimization objective, enabling the discovery of architectures that achieve strong performance while minimizing computational cost. Experiments on five benchmark datasets (MNIST, Digits, Wine, Breast Cancer, and Iris) show that quantum FLOPs dominate accuracy improvements, while classical FLOPs remain largely fixed. Pareto-optimal solutions reveal that competitive accuracy can often be achieved with significantly reduced computational cost compared to FLOPs-agnostic baselines. Our results establish FLOPs-awareness as a practical criterion for HQNN design in the NISQ era and as a scalable principle for future HQNN systems.
Post-Variational Quantum Neural Networks (PVQNNs) eliminate the need for quantum parameter optimization by employing fixed, non-parametric circuits and shifting all learning to the classical domain, thereby avoiding barren plateaus entirely. However, the performance of PV-QNNs critically depends on the choice of ansatz architecture. In this work, we propose an automatic ansatz optimization framework based on Genetic Algorithms (GAs) that searches for optimal circuit ensembles under discrete constraints of topology, depth, and gate types. Quantum features are extracted efficiently via classical shadows and processed by a shallow classical classifier, preserving the post-variational property. Experimental results on the Fashion-MNIST (0 vs 4) classification task demonstrate that GA-selected ensembles outperform all baselines, achieving 89.0 % accuracy with 8 qubits, exceeding random and manually designed templates by3−5%. Moreover, GA ensembles achieve up to1.9×higher accuracy per circuit and2×higher accuracy per twoqubit gate, while maintaining shallower circuits (6.0 depth vs. 8.6). These results indicate that evolutionary search effectively identifies compact, information-rich quantum circuits, enabling scalable and resource-efficient post-variational learning.