Abstract:Quantum Machine Learning (QML) has surfaced as a pioneering framework addressing sequential control tasks and time-series modeling. It has demonstrated empirical quantum advantages notably within domains such as Reinforcement Learning (RL) and time-series prediction. A significant advancement lies in Quantum Recurrent Neural Networks (QRNNs), specifically tailored for memory-intensive tasks encompassing partially observable environments and non-linear time-series prediction. Nevertheless, QRNN-based models encounter challenges, notably prolonged training duration stemming from the necessity to compute quantum gradients using backpropagation-through-time (BPTT). This predicament exacerbates when executing the complete model on quantum devices, primarily due to the substantial demand for circuit evaluation arising from the parameter-shift rule. This paper introduces the Quantum Fast Weight Programmers (QFWP) as a solution to the temporal or sequential learning challenge. The QFWP leverages a classical neural network (referred to as the 'slow programmer') functioning as a quantum programmer to swiftly modify the parameters of a variational quantum circuit (termed the 'fast programmer'). Instead of completely overwriting the fast programmer at each time-step, the slow programmer generates parameter changes or updates for the quantum circuit parameters. This approach enables the fast programmer to incorporate past observations or information. Notably, the proposed QFWP model achieves learning of temporal dependencies without necessitating the use of quantum recurrent neural networks. Numerical simulations conducted in this study showcase the efficacy of the proposed QFWP model in both time-series prediction and RL tasks. The model exhibits performance levels either comparable to or surpassing those achieved by QLSTM-based models.

A Quantum Circuit-Based Compression Perspective for Parameter-Efficient Learning

Learning Parameterized Quantum Circuits with Quantum Gradient

Parameterized Quantum Circuits with Quantum Kernels for Machine Learning: A Hybrid Quantum-Classical Approach

Training Classical Neural Networks by Quantum Machine Learning

Quantum-Train: Rethinking Hybrid Quantum-Classical Machine Learning in the Model Compression Perspective

Learning to Program Variational Quantum Circuits with Fast Weights

Diffusion-Inspired Quantum Noise Mitigation in Parameterized Quantum Circuits

On Optimizing Hyperparameters for Quantum Neural Networks

Graph Learning for Parameter Prediction of Quantum Approximate Optimization Algorithm

Programming Variational Quantum Circuits with Quantum-Train Agent

L2O-$g^{\dagger}$: Learning to Optimize Parameterized Quantum Circuits with Fubini-Study Metric Tensor

A Study on Optimization Techniques for Variational Quantum Circuits in Reinforcement Learning

Non-parametric Greedy Optimization of Parametric Quantum Circuits

A Parameter-Efficient Quantum Anomaly Detection Method on a Superconducting Quantum Processor

Finding Optimal Training Parameters for Quantum Generative Adversarial Networks

Quantum generative adversarial learning for simultaneous multiparameter estimation

Non-asymptotic Approximation Error Bounds of Parameterized Quantum Circuits

Federated Quantum-Train with Batched Parameter Generation

Improving Parameter Training for VQEs by Sequential Hamiltonian Assembly

Quantum-inspired activation functions and quantum Chebyshev-polynomial network

Pulse-efficient quantum machine learning