Abstract Variational quantum algorithms represent a powerful approach for solving optimization problems on noisy quantum computers, with a broad spectrum of potential applications ranging from chemistry to machine learning. However, their performances in practical implementations crucially depend on the effectiveness of quantum circuit training, which can be severely limited by phenomena such as barren plateaus. While, in general, dissipation is detrimental for quantum algorithms, and noise itself can actually induce barren plateaus, here we describe how the inclusion of properly engineered Markovian losses after each unitary quantum circuit layer allows for the trainability of quantum models. We identify the required form of the dissipation processes and establish that their optimization is efficient. We benchmark the generality of our proposal in both a synthetic and a practical quantum chemistry example, demonstrating its effectiveness and potential impact across different domains.
Abstract Variational quantum algorithms represent a powerful approach for solving optimization problems on noisy quantum computers, with a broad spectrum of potential applications ranging from chemistry to machine learning. However, their performances in practical implementations crucially depend on the effectiveness of quantum circuit training, which can be severely limited by phenomena such as barren plateaus. While, in general, dissipation is detrimental for quantum algorithms, and noise itself can actually induce barren plateaus, here we describe how the inclusion of properly engineered Markovian losses after each unitary quantum circuit layer allows for the trainability of quantum models. We identify the required form of the dissipation processes and establish that their optimization is efficient. We benchmark the generality of our proposal in both a synthetic and a practical quantum chemistry example, demonstrating its effectiveness and potential impact across different domains.
Quantum information processing holds immense promise, but its practical implementation is fundamentally challenged by dissipation, the irreversible loss of information and energy to the environment. The conventional approach treats dissipation as an obstacle to overcome, typically through resource-intensive quantum error correction. Such schemes impose overheads of orders of magnitude in terms of both qubits and operations. This creates significant space, control and time costs, making near-term practical implementation highly challenging. This thesis adopts a different viewpoint, harnessing dissipation for quantum algorithms. We explore the design, application, and conceptual advantages of dissipative quantum algorithms. We establish the foundational tools for this approach by demonstrating how to engineer a universal set of nonunitary operations using intrinsic, controllable dissipation. We present a hardware-efficient method, experimentally realized in a trapped-ion system, for implementing any two-qubit Markovian quantum channel without requiring ancilla qubits or measurements. This provides a practical and versatile toolbox for constructing dissipative algorithms on near-term quantum devices. Building on this, we apply these tools to an inherently dissipative task: the preparation of thermal states. We introduce a variational quantum-classical algorithm that integrates engineered nonunitary operations with unitary gates. By leveraging weak symmetries of the target Hamiltonian, we design efficient ansätze that significantly improve the fidelity of state preparation. Finally, we address one of the most significant barriers to scaling variational quantum algorithms: the prevalence of barren plateaus, where optimization gradients vanish exponentially with system size. We prove that dissipative algorithms, by incorporating periodic qubit resets, can mitigate both noise-induced and unitary barren plateaus. This mechanism allows for a class of scalable computations that are intractable for purely unitary approaches in the presence of noise. Together, these results demonstrate that dissipation is not just a source of error, but a powerful resource. By systematically developing the necessary operations, applying them to relevant problems, and identifying fundamental scalability advantages, this work showcases dissipative quantum computing as a compelling avenue in quantum information processing.
Abstract In recent years, variational quantum circuits (VQCs) have been widely explored to advance quantum circuits against classic models on various domains, such as quantum chemistry and quantum machine learning. Similar to classic machine-learning models, VQCs can be trained through various optimization approaches, such as gradient-based or gradient-free methods. However, when employing gradient-based methods, the gradient variance of VQCs may dramatically vanish as the number of qubits or layers increases. This issue, a.k.a. barren plateaus (BPs), seriously hinders the scaling of VQCs on large datasets. To mitigate the barren plateaus, extensive efforts have been devoted to tackling this issue through diverse strategies. In this survey, we conduct a systematic literature review of recent works from both investigation and mitigation perspectives. Furthermore, we propose a new taxonomy to categorize most existing mitigation strategies into five groups and introduce them in detail. Also, we compare the concurrent survey papers about BPs. Finally, we provide insightful discussion on future directions for BPs.
Abstract Combining classical optimization with parameterized quantum circuit evaluation, variational quantum algorithms (VQAs) are among the most promising algorithms in near-term quantum computing. However, the training of variational quantum circuits (VQCs) is susceptible to a phenomenon known as barren plateaus (BPs). Various approaches have been proposed to mitigate this issue, such as using neural networks to generate VQC parameters. In this paper, we improve the NN-based BP mitigation approach by refining the neural network architecture and extend its applicability to a more generalized scenario that includes random quantum inputs and VQC structures. We evaluate the effectiveness of the approach by comparing the convergence speed before and after it is utilized. Furthermore, we give an explanation for the effectiveness of the NN-based BP mitigation approach by utilizing a loss landscape visualization technique and the expressibility metric of VQC. The smoothness of the loss landscape offers an intuitive insight into the approach’s utility, while the reduction in expressibility accounts for the enhanced trainability. Our research highlights the universal applicability of the NN-based BP mitigation approach, underscoring its potential to drive progress in the development of VQAs across diverse domains.
Variational quantum algorithms (VQAs) are a promising paradigm for quantum advantage, yet their trainability is severely hampered by barren plateaus (BPs). Several recent works have proposed dynamic parameterized quantum circuits (DPQCs), which interleave unitary layers with parameterized CPTP maps, such as engineered dissipation, feedforward gadgets, and periodic resets, as a possible strategy for mitigating BPs. We unify this class of circuits into a formalization for DPQCs.We identify constraints on the nature and the structure of DPQCs if they are to prevent a significant number of parameters from becoming untrainable. Using purification and Pauli-path analysis, we further identify a mechanism by which the cost function can remain anti-concentrated even when many parameters remain untrainable. Our analysis reveals ways to design DPQCs that do not have an exponentially concentrated cost function, and our results suggest that BP mitigation via DPQCs is at least as hard as designing BP-free unitaries.
In the search for quantum advantage with near-term quantum devices, navigating the optimization landscape is significantly hampered by the barren plateaus phenomenon. This study presents a strategy to overcome this obstacle without changing the quantum circuit architecture. We propose incorporating auxiliary control qubits to shift the circuit from a unitary 2-design to a unitary 1-design, mitigating the prevalence of barren plateaus. We then remove these auxiliary qubits to return to the original circuit structure while preserving the unitary 1-design properties. Our numerical simulation suggests that the proposed structure effectively mitigates the barren plateaus phenomenon. A significant finding is that the gradient of ${\ensuremath{\theta}}_{1,1}$, the first parameter in the quantum circuit, displays a broader distribution as the number of qubits and layers increases. This suggests a higher probability of obtaining effective gradients. This stability is critical for the efficient training of quantum circuits, especially for larger and more complex systems. The results of this study represent a significant advance in the optimization of quantum circuits and offer a promising avenue for the scalable and practical implementation of quantum computing technologies. This approach opens up new opportunities in quantum learning and other applications that require robust quantum computing power.
Variational quantum algorithms (VQAs) promise efficient use of near-term quantum computers. However, training VQAs often requires an extensive amount of time and suffers from the barren plateau problem where the magnitude of the gradients vanishes with increasing number of qubits. Here, we show how to optimally train VQAs for learning quantum states. Parameterized quantum circuits can form Gaussian kernels, which we use to derive adaptive learning rates for gradient ascent. We introduce the generalized quantum natural gradient that features stability and optimized movement in parameter space. Both methods together outperform other optimization routines in training VQAs. Our methods also excel at numerically optimizing driving protocols for quantum control problems. The gradients of the VQA do not vanish when the fidelity between the initial state and the state to be learned is bounded from below. We identify a VQA for quantum simulation with such a constraint that thus can be trained free of barren plateaus. Finally, we propose the application of Gaussian kernels for quantum machine learning.