A Review of Barren Plateaus in Variational Quantum Computing

Martin Larocca,Supanut Thanasilp,Samson Wang,Kunal Sharma,Jacob Biamonte,Patrick J. Coles,Lukasz Cincio,Jarrod R. McClean,Zoë Holmes,M. Cerezo

2024-05-02

Abstract:Variational quantum computing offers a flexible computational paradigm with applications in diverse areas. However, a key obstacle to realizing their potential is the Barren Plateau (BP) phenomenon. When a model exhibits a BP, its parameter optimization landscape becomes exponentially flat and featureless as the problem size increases. Importantly, all the moving pieces of an algorithm -- choices of ansatz, initial state, observable, loss function and hardware noise -- can lead to BPs when ill-suited. Due to the significant impact of BPs on trainability, researchers have dedicated considerable effort to develop theoretical and heuristic methods to understand and mitigate their effects. As a result, the study of BPs has become a thriving area of research, influencing and cross-fertilizing other fields such as quantum optimal control, tensor networks, and learning theory. This article provides a comprehensive review of the current understanding of the BP phenomenon.

Quantum Physics,Machine Learning

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the "Barren Plateau" (BP) phenomenon in Variational Quantum Computing (VQC). The BP phenomenon means that during the parameter optimization process, as the system scale increases, the gradient of the loss function or more generally the loss difference approaches zero exponentially, resulting in an extremely flat and featureless optimization landscape. This phenomenon seriously affects the training efficiency and scalability of variational quantum algorithms.

Specifically, the paper focuses on the following points:

1. **Definition and Identification of BP**: The paper first defines different types of BP, including probability - concentrated BP and deterministic BP, and discusses how to identify these phenomena by analyzing the variance of the loss function and its gradient.

2. **Origin of BP**: The paper explores the root cause of the BP phenomenon, pointing out that it is the "curse of dimensionality" caused by the exponential growth of the dimension of the Hilbert space in variational quantum computing with the system scale. This increase in dimension makes the inner product between the state after most parameter initializations and the target measurement operator very small, thus leading to the BP phenomenon.

3. **Influencing Factors**: The paper analyzes in detail various factors that may lead to BP, including the design of the quantum circuit (such as the expressive power of the circuit), the choice of the initial state, the setting of the measurement operator, and the influence of hardware noise, etc.

4. **Mitigation Strategies**: The paper also reviews some theoretical and heuristic methods proposed in current research, aiming to understand and alleviate the impact of the BP phenomenon on the training process, for example, by introducing appropriate inductive biases to design quantum circuits, or adopting specific optimization strategies and techniques.

In summary, this paper provides a comprehensive review, aiming to help researchers better understand the nature of the BP phenomenon and explore effective solutions to promote the development and application of variational quantum computing.