Maximising Quantum-Computing Expressive Power through Randomised Circuits

Yingli Yang,Zongkang Zhang,Anbang Wang,Xiaosi Xu,Xiaoting Wang,Ying Li
DOI: https://doi.org/10.1103/PhysRevResearch.6.023098
2023-12-04
Abstract:In the noisy intermediate-scale quantum era, variational quantum algorithms (VQAs) have emerged as a promising avenue to obtain quantum advantage. However, the success of VQAs depends on the expressive power of parameterised quantum circuits, which is constrained by the limited gate number and the presence of barren plateaus. In this work, we propose and numerically demonstrate a novel approach for VQAs, utilizing randomised quantum circuits to generate the variational wavefunction. We parameterize the distribution function of these random circuits using artificial neural networks and optimize it to find the solution. This random-circuit approach presents a trade-off between the expressive power of the variational wavefunction and time cost, in terms of the sampling cost of quantum circuits. Given a fixed gate number, we can systematically increase the expressive power by extending the quantum-computing time. With a sufficiently large permissible time cost, the variational wavefunction can approximate any quantum state with arbitrary accuracy. Furthermore, we establish explicit relationships between expressive power, time cost, and gate number for variational quantum eigensolvers. These results highlight the promising potential of the random-circuit approach in achieving a high expressive power in quantum computing.
Quantum Physics,Machine Learning
What problem does this paper attempt to address?
The problem that this paper attempts to solve is how to maximize the expressive power of variational quantum algorithms (VQAs) through random circuits in quantum computing. Specifically, the paper focuses on the fact that in the noisy intermediate - scale quantum (NISQ) era, due to the limited number of quantum gates and the existence of barren plateaus, the expressive power of variational quantum circuits is restricted. To overcome these limitations, the paper proposes a new method, that is, using randomized circuits to generate variational wave functions, and parameterizing the distribution functions of these randomized circuits through artificial neural networks (ANNs), thereby optimizing the variational wave functions to solve the problem. ### Main Problems and Solutions 1. **Limited Expressive Power**: - **Problem**: On NISQ devices, the expressive power of variational quantum circuits is limited by the number of quantum gates, which causes the variational wave function to represent only a subset of all possible quantum states. - **Solution**: The paper proposes using randomized circuits to generate variational wave functions and parameterizing the distribution functions of these randomized circuits through ANN, thereby systematically increasing the expressive power while fixing the number of quantum gates. 2. **The Problem of Vanishing Gradients during the Training Process**: - **Problem**: High - expressive - power circuits are usually similar to unitary t - designs, which will lead to the problem of vanishing gradients during the training process. - **Solution**: By introducing randomized circuits and ANN - parameterized distribution functions, the paper provides a new method to optimize the variational wave function, thereby avoiding the problem of vanishing gradients. ### Main Contributions of the Paper 1. **Random Circuit Method**: - A new method of using randomized circuits to generate variational wave functions is proposed. This method can improve the expressive power by increasing the time cost while fixing the number of quantum gates. - The optimization of variational wave functions is achieved by parameterizing the distribution functions and phase functions of randomized circuits through ANN. 2. **Theoretical Analysis**: - It is proved that in the low - time - cost limit, the expressive power of the random circuit method is not lower than that of the deterministic circuit method. - It is proved that in the high - time - cost limit, the random circuit method can approximate any quantum state to any precision. - The trade - off relationship between the expressive power and the time cost, as well as the specific upper - bound error analysis, are given. 3. **Numerical Experiments**: - The performance of the random circuit method in solving the ground - state problem of the antiferromagnetic Heisenberg model is demonstrated through numerical simulation, and the trade - off relationship between the expressive power and the time cost is verified. ### Conclusion The paper proposes a method of using randomized circuits and ANN - parameterized distribution functions, which can systematically improve the expressive power of variational quantum algorithms while fixing the number of quantum gates. This method not only solves the problems of limited expressive power and vanishing gradients, but also provides new ideas for future quantum computing applications.