Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

IEEE Transactions on Pattern Analysis and Machine Intelligence

Quantum Gated Recurrent Neural Networks

Yanan Li, Zhimin Wang, Ruipeng Xing, Changheng Shao, Shangshang Shi, Jiaxin Li, Guoqiang Zhong, Yongjian Gu

Published Dec 18, 2024• 19 citations
Publisher

Abstract

The exploration of quantum advantages with Quantum Neural Networks (QNNs) is an exciting endeavor. Recurrent neural networks, the widely used framework in deep learning, suffer from the gradient vanishing and exploding problem, which limits their ability to learn long-term dependencies. To address this challenge, in this work, we develop the sequential model of Quantum Gated Recurrent Neural Networks (QGRNNs). This model naturally integrates the gating mechanism into the framework of the variational ansatz circuit of QNNs, enabling efficient execution on near-term quantum devices. We present rigorous proof that QGRNNs can preserve the gradient norm of long-term interactions throughout the recurrent network, enabling efficient learning of long-term dependencies. Meanwhile, the architectural features of QGRNNs can effectively mitigate the barren plateau phenomenon. The effectiveness of QGRNNs in sequential learning is convincingly demonstrated through various typical tasks, including solving the adding problem, learning gene regulatory networks, and predicting stock prices. The hardware-efficient architecture and superior performance of our QGRNNs indicate their promising potential for finding quantum advantageous applications in the near term.

DOI
10.1109/tpami.2024.3519605

Related papers

Research and development of a quantum recurrent neural network for learning

S.M. Gushanskiy, Viktor Potapov

Quantum neural network (QNN) is one of the promising areas in which noisy medium-scale quantum devices (MCD) can find advantageous applications in the short term compared to classical resources. Recurrent neural networks are the most fundamental networks for sequential learning, but there is still no canonical model of a quantum recurrent neural network (QRNN), which certainly limits research in the field of quantum deep learning. In the present paper, a new type of QRNS is proposed that would be a good candidate as a canonical QRNN model, where quantum recurrent blocks (QRB) are built in a hardware-efficient way, and QRNN are constructed by stacking QRB in a checkerboard pattern, which can significantly reduce the algorithm’s requirements with respect to coherent time of quantum devices. That is, QRNN is much more accessible on MCD devices. Numerical experiments show that our QRNN provides much better performance in prediction (classification) accuracy compared to classical RNN and state-of-the-art QNN models for sequential training and can predict changing details of time series data.

Informatization and communication•Jan 1, 2023•0 citations

A fully quantum-native recurrent neural network for end-to-end sequential learning on NISQ hardware

Rui Huang, Haibo Yi

Abstract Modeling temporal dependencies within quantum systems remains a key challenge for quantum machine learning. Current quantum neural networks largely depend on classical recurrent modules, which introduce optimization bottlenecks and coherence loss during sequence processing. To address this, we propose a fully quantum-native recurrent neural network (QRNN) for end-to-end sequential learning. Our architecture captures temporal dynamics through coherent unitary operations, avoiding external classical recurrence and mid-circuit state collapse. The proposed model was benchmarked against a classical recurrent neural network (RNN), quantum tensor networks, and state-of-the-art hybrid QRNN using the Origin Quantum superconducting platform. Experimental evaluations on nonlinear functions, complex dynamical systems, and meteorological datasets demonstrate that our quantum-native QRNN achieves improved prediction accuracy and exhibits enhanced capability in forecasting complex temporal features. For instance, on the relative humidity forecasting task, the proposed QRNN achieves a 73.5% reduction in Root mean square error compared to the hybrid VQRNN. This work establishes a robust framework for preserving coherence in pure quantum sequential processing, offering a scalable path for time-series forecasting on near-term quantum hardware.

Machine Learning Science and Technology•Apr 15, 2026•3 citations

Reservoir Computing via Quantum Recurrent Neural Networks

Samuel Yen-Chi Chen, Daniel Fry, Deshmukh, Amol +2 more

Recent developments in quantum computing and machine learning have propelled the interdisciplinary study of quantum machine learning. Sequential modeling is an important task with high scientific and commercial value. Existing VQC or QNN-based methods require significant computational resources to perform the gradient-based optimization of a larger number of quantum circuit parameters. The major drawback is that such quantum gradient calculation requires a large amount of circuit evaluation, posing challenges in current near-term quantum hardware and simulation software. In this work, we approach sequential modeling by applying a reservoir computing (RC) framework to quantum recurrent neural networks (QRNN-RC) that are based on classical RNN, LSTM and GRU. The main idea to this RC approach is that the QRNN with randomly initialized weights is treated as a dynamical system and only the final classical linear layer is trained. Our numerical simulations show that the QRNN-RC can reach results comparable to fully trained QRNN models for several function approximation and time series prediction tasks. Since the QRNN training complexity is significantly reduced, the proposed model trains notably faster. In this work we also compare to corresponding classical RNN-based RC implementations and show that the quantum version learns faster by requiring fewer training epochs in most cases. Our results demonstrate a new possibility to utilize quantum neural network for sequential modeling with greater quantum hardware efficiency, an important design consideration for noisy intermediate-scale quantum (NISQ) computers.

arXiv:2211.02612•Nov 4, 2022•32 citations

Quantum gated recurrent neural networks for continuous dynamic temporal decision-making

JiaZhao Shen, Zijun Guo, Wei Ding +2 more

Abstract Performance breakthroughs in dynamic temporal decision-making tasks have long been a core research objective in reinforcement learning. Yet traditional policy networks, including long short-term memory (LSTM) networks, gated recurrent unit (GRU) and feedforward neural network (FNN), have long been hampered by inherent bottlenecks: low reward efficiency and training instability, which renders them less capable of meeting the high-precision decision-making demands of complex scenarios. This paper introduces a novel policy network, quantum gated recurrent neural network (QGRNN), which incorporates quantum superposition encoding and cyclic entanglement structures to achieve more efficient state-space exploration. Furthermore, the network employs a quantum–classical hybrid residual design that effectively integrates quantum measurement outcomes with classical inputs, thereby strengthening the representational capacity for temporal decision-making features. Trained over 800 episodes via proximal policy optimization in CartPole-v1, the QGRNN achieves an average reward of 380.3 ± 128.6 in the last 100 episodes-significantly outperforming GRU (275.9 ± 140.7), LSTM (239.7 ± 131.8), and FNN (219.2 ± 109.5). As a 4-qubit proof-of-concept, it achieves 45% high-performance episodes (rewards > 450) and demonstrates low-resource, high-performance characteristics for temporal decision-making.

Physica Scripta•Aug 14, 2026•0 citations

Quantum Neural Networks for Pattern Recognition and Prediction

Oscar Muller, Elena Schmidt, Anna Horvath

Quantum neural networks (QNNs) -- parameterised quantum circuits trained to perform pattern recognition and predictive tasks -- represent a prominent but contested direction in quantum machine learning, offering potential parameter efficiency and expressibility benefits over classical neural networks for specific problem structures while suffering from barren plateau trainability limitations at scale. The 2024 issue of this journal established SNAD LSAD-structured ansatz (Silva et al., 2024) as the critical enabler of scalable QNN training, and QHDC (Bianchi et al., 2024) showed that QNN advantage is dataset-dependent. This paper advances the QNN field with the Quantum Neural Network Design and Evaluation (QNNDE) framework, a systematic methodology for QNN architecture design, training optimisation, and performance evaluation across four pattern recognition and prediction task classes: temporal sequence modelling (quantum recurrent circuits), spatial pattern recognition (quantum convolutional circuits), generative modelling (quantum generative adversarial networks), and time-series regression (quantum reservoir computing). QNNDE is evaluated on eight benchmark tasks spanning physics simulation, molecular property prediction, quantum state tomography, financial time-series, medical signal classification, and quantum circuit synthesis on IBM Quantum Eagle (127-qubit) and IonQ Forte (35-qubit). Quantum recurrent circuits achieve 12.4pp accuracy improvement over classical LSTMs for quantum state evolution prediction. Quantum convolutional circuits match classical CNNs on image-derived molecular patterns at 68.4% parameter reduction. Quantum GANs generate molecular conformations with 8.4% lower FID than classical GAN at equal network size. Quantum reservoir computing outperforms classical echo state networks by 18.4% on quantum-process time-series prediction.

Quantum Frontiers Journal•Mar 28, 2025•0 citations

Quantum-Enhanced Recurrent Neural Networks via Variational Quantum Gating for Battery State of Health Prediction

Yin Xu, Qinglin Liu, Li Gao +1 more

Accurate state-of-health (SOH) estimation for lithium-ion batteries remains a challenging problem due to complex electrochemical degradation mechanisms and long-range temporal dependencies. In this work, we propose a quantum-enhanced recurrent framework, termed QLSTM, in which variational quantum circuits are directly embedded into the gating mechanisms of long short-term memory networks. By replacing classical affine transformations with parameterized unitary operations, the proposed model introduces structured nonlinear transformations into the recurrent state-transition process. Extensive experiments on multiple benchmark battery datasets demonstrate that QLSTM consistently outperforms classical sequence models in both predictive accuracy and robustness, achieving significant reductions in mean absolute error (MAE), with improvements on the order of 20% compared with classical LSTM baselines. Ablation studies further confirm that these improvements arise primarily from quantum-enhanced gating rather than input-level transformations. Additional analyses on qubit scaling and noise robustness reveal that model performance is governed by a balance between expressive capacity and trainability. These results provide empirical evidence that embedding quantum computational primitives within recurrent architectures offers a structurally grounded approach to improving sequence modeling capability. The proposed framework establishes a new design paradigm for integrating quantum operators into temporal learning models, with potential applications in complex dynamical system prediction tasks.

quant-pharXiv:2604.20438•Apr 23, 2026•0 citations
The exploration of quantum advantages with Quantum Neural Networks (QNNs) is an exciting endeavor. Recurrent neural networks, the widely used framework in deep learning, suffer from the gradient vanishing and exploding problem, which limits their ability to learn long-term dependencies. To address this challenge, in this work, we develop the sequential model of Quantum Gated Recurrent Neural Networks (QGRNNs). This model naturally integrates the gating mechanism into the framework of the variational ansatz circuit of QNNs, enabling efficient execution on near-term quantum devices. We present rigorous proof that QGRNNs can preserve the gradient norm of long-term interactions throughout the recurrent network, enabling efficient learning of long-term dependencies. Meanwhile, the architectural features of QGRNNs can effectively mitigate the barren plateau phenomenon. The effectiveness of QGRNNs in sequential learning is convincingly demonstrated through various typical tasks, including solving the adding problem, learning gene regulatory networks, and predicting stock prices. The hardware-efficient architecture and superior performance of our QGRNNs indicate their promising potential for finding quantum advantageous applications in the near term.
Quantum neural network (QNN) is one of the promising areas in which noisy medium-scale quantum devices (MCD) can find advantageous applications in the short term compared to classical resources. Recurrent neural networks are the most fundamental networks for sequential learning, but there is still no canonical model of a quantum recurrent neural network (QRNN), which certainly limits research in the field of quantum deep learning. In the present paper, a new type of QRNS is proposed that would be a good candidate as a canonical QRNN model, where quantum recurrent blocks (QRB) are built in a hardware-efficient way, and QRNN are constructed by stacking QRB in a checkerboard pattern, which can significantly reduce the algorithm’s requirements with respect to coherent time of quantum devices. That is, QRNN is much more accessible on MCD devices. Numerical experiments show that our QRNN provides much better performance in prediction (classification) accuracy compared to classical RNN and state-of-the-art QNN models for sequential training and can predict changing details of time series data.
Abstract Modeling temporal dependencies within quantum systems remains a key challenge for quantum machine learning. Current quantum neural networks largely depend on classical recurrent modules, which introduce optimization bottlenecks and coherence loss during sequence processing. To address this, we propose a fully quantum-native recurrent neural network (QRNN) for end-to-end sequential learning. Our architecture captures temporal dynamics through coherent unitary operations, avoiding external classical recurrence and mid-circuit state collapse. The proposed model was benchmarked against a classical recurrent neural network (RNN), quantum tensor networks, and state-of-the-art hybrid QRNN using the Origin Quantum superconducting platform. Experimental evaluations on nonlinear functions, complex dynamical systems, and meteorological datasets demonstrate that our quantum-native QRNN achieves improved prediction accuracy and exhibits enhanced capability in forecasting complex temporal features. For instance, on the relative humidity forecasting task, the proposed QRNN achieves a 73.5% reduction in Root mean square error compared to the hybrid VQRNN. This work establishes a robust framework for preserving coherence in pure quantum sequential processing, offering a scalable path for time-series forecasting on near-term quantum hardware.
Recent developments in quantum computing and machine learning have propelled the interdisciplinary study of quantum machine learning. Sequential modeling is an important task with high scientific and commercial value. Existing VQC or QNN-based methods require significant computational resources to perform the gradient-based optimization of a larger number of quantum circuit parameters. The major drawback is that such quantum gradient calculation requires a large amount of circuit evaluation, posing challenges in current near-term quantum hardware and simulation software. In this work, we approach sequential modeling by applying a reservoir computing (RC) framework to quantum recurrent neural networks (QRNN-RC) that are based on classical RNN, LSTM and GRU. The main idea to this RC approach is that the QRNN with randomly initialized weights is treated as a dynamical system and only the final classical linear layer is trained. Our numerical simulations show that the QRNN-RC can reach results comparable to fully trained QRNN models for several function approximation and time series prediction tasks. Since the QRNN training complexity is significantly reduced, the proposed model trains notably faster. In this work we also compare to corresponding classical RNN-based RC implementations and show that the quantum version learns faster by requiring fewer training epochs in most cases. Our results demonstrate a new possibility to utilize quantum neural network for sequential modeling with greater quantum hardware efficiency, an important design consideration for noisy intermediate-scale quantum (NISQ) computers.
Abstract Performance breakthroughs in dynamic temporal decision-making tasks have long been a core research objective in reinforcement learning. Yet traditional policy networks, including long short-term memory (LSTM) networks, gated recurrent unit (GRU) and feedforward neural network (FNN), have long been hampered by inherent bottlenecks: low reward efficiency and training instability, which renders them less capable of meeting the high-precision decision-making demands of complex scenarios. This paper introduces a novel policy network, quantum gated recurrent neural network (QGRNN), which incorporates quantum superposition encoding and cyclic entanglement structures to achieve more efficient state-space exploration. Furthermore, the network employs a quantum–classical hybrid residual design that effectively integrates quantum measurement outcomes with classical inputs, thereby strengthening the representational capacity for temporal decision-making features. Trained over 800 episodes via proximal policy optimization in CartPole-v1, the QGRNN achieves an average reward of 380.3 ± 128.6 in the last 100 episodes-significantly outperforming GRU (275.9 ± 140.7), LSTM (239.7 ± 131.8), and FNN (219.2 ± 109.5). As a 4-qubit proof-of-concept, it achieves 45% high-performance episodes (rewards > 450) and demonstrates low-resource, high-performance characteristics for temporal decision-making.
Quantum neural networks (QNNs) -- parameterised quantum circuits trained to perform pattern recognition and predictive tasks -- represent a prominent but contested direction in quantum machine learning, offering potential parameter efficiency and expressibility benefits over classical neural networks for specific problem structures while suffering from barren plateau trainability limitations at scale. The 2024 issue of this journal established SNAD LSAD-structured ansatz (Silva et al., 2024) as the critical enabler of scalable QNN training, and QHDC (Bianchi et al., 2024) showed that QNN advantage is dataset-dependent. This paper advances the QNN field with the Quantum Neural Network Design and Evaluation (QNNDE) framework, a systematic methodology for QNN architecture design, training optimisation, and performance evaluation across four pattern recognition and prediction task classes: temporal sequence modelling (quantum recurrent circuits), spatial pattern recognition (quantum convolutional circuits), generative modelling (quantum generative adversarial networks), and time-series regression (quantum reservoir computing). QNNDE is evaluated on eight benchmark tasks spanning physics simulation, molecular property prediction, quantum state tomography, financial time-series, medical signal classification, and quantum circuit synthesis on IBM Quantum Eagle (127-qubit) and IonQ Forte (35-qubit). Quantum recurrent circuits achieve 12.4pp accuracy improvement over classical LSTMs for quantum state evolution prediction. Quantum convolutional circuits match classical CNNs on image-derived molecular patterns at 68.4% parameter reduction. Quantum GANs generate molecular conformations with 8.4% lower FID than classical GAN at equal network size. Quantum reservoir computing outperforms classical echo state networks by 18.4% on quantum-process time-series prediction.
Accurate state-of-health (SOH) estimation for lithium-ion batteries remains a challenging problem due to complex electrochemical degradation mechanisms and long-range temporal dependencies. In this work, we propose a quantum-enhanced recurrent framework, termed QLSTM, in which variational quantum circuits are directly embedded into the gating mechanisms of long short-term memory networks. By replacing classical affine transformations with parameterized unitary operations, the proposed model introduces structured nonlinear transformations into the recurrent state-transition process. Extensive experiments on multiple benchmark battery datasets demonstrate that QLSTM consistently outperforms classical sequence models in both predictive accuracy and robustness, achieving significant reductions in mean absolute error (MAE), with improvements on the order of 20% compared with classical LSTM baselines. Ablation studies further confirm that these improvements arise primarily from quantum-enhanced gating rather than input-level transformations. Additional analyses on qubit scaling and noise robustness reveal that model performance is governed by a balance between expressive capacity and trainability. These results provide empirical evidence that embedding quantum computational primitives within recurrent architectures offers a structurally grounded approach to improving sequence modeling capability. The proposed framework establishes a new design paradigm for integrating quantum operators into temporal learning models, with potential applications in complex dynamical system prediction tasks.