Quantum reservoir computing (QRC) harnesses driven quantum dynamics for time-series processing, yet the mechanisms behind the differing performance levels across its many implementations remain unclear. We show that apparently unrelated approaches-including memory restriction, weak measurements, operation near the edge of quantum chaos, and dissipative dynamics-are in fact governed by the same underlying principle, namely a tunable balance between memory retention and nonlinear response. Using the information processing capacity, a dynamical measure from nonlinear systems theory, we place these behaviors in a unified framework and identify the regimes in which quantum reservoirs surpass the standard protocol. Our results reveal a fundamental connection between memory and nonlinear response. This provides a general design principle for enhanced information processing and enables systematic analysis and optimization inspired by classical dynamical quantifiers.
Quantum reservoir computing (QRC) harnesses driven quantum dynamics for time-series processing, yet the mechanisms behind the differing performance levels across its many implementations remain unclear. We show that apparently unrelated approaches-including memory restriction, weak measurements, operation near the edge of quantum chaos, and dissipative dynamics-are in fact governed by the same underlying principle, namely a tunable balance between memory retention and nonlinear response. Using the information processing capacity, a dynamical measure from nonlinear systems theory, we place these behaviors in a unified framework and identify the regimes in which quantum reservoirs surpass the standard protocol. Our results reveal a fundamental connection between memory and nonlinear response. This provides a general design principle for enhanced information processing and enables systematic analysis and optimization inspired by classical dynamical quantifiers.
Numerical studies suggest that quantum reservoir computing performs optimally near the edge of chaos, a behavior commonly attributed to a balance between memory retention and nonlinear information processing. Here we develop an explicit quantum reservoir model and rigorously analyze its chaotic regime and integrable point. In the chaotic regime, we show that Gaussian-unitary-ensemble statistics lead to rapid memory loss: learning performance decays exponentially with the storage time required by a temporal task. This decay limits the ability of chaotic reservoirs to learn long-range temporal dependencies. At the integrable point, described by a free-fermion model, we prove that amplitude encoding and a linear readout of fermionic two-point correlations cannot learn a representative class of nonlinear tasks coupling present and past inputs. Introducing interactions alleviates this obstruction. By identifying distinct limitations on memory retention in the chaotic regime and nonlinear processing at the integrable point, our analysis provides a theoretical basis for the observed edge-of-chaos advantage and guides the design of quantum reservoirs for temporal learning.
Quantum reservoir computing (QRC) is a highly promising computational paradigm that leverages quantum systems as a computational resource for nonlinear information processing. While its application to time-series analysis is eagerly anticipated, prevailing approaches suffer from the collapse of the quantum state upon measurement, resulting in the erasure of temporal input memories. Neither repeated initializations nor weak measurements offer a fundamental solution, as the former escalates the time complexity while the latter restricts the information extraction from the Hilbert space. To address this issue, we propose the feedback-driven QRC framework. This methodology employs projective measurements on all qubits for unrestricted access to the quantum state, with the measurement outcomes subsequently fed back into the reservoir to restore the memory of prior inputs. We demonstrate that our QRC successfully acquires the fading-memory property through the feedback connections, a critical aspect in time-series processing. Notably, analysis of measurement trajectories reveals three distinct phases depending on the feedback strength, with the memory performance maximized at the edge of chaos. We also evaluate the predictive capabilities of our QRC, demonstrating its suitability for forecasting signals originating from quantum spin systems. Published by the American Physical Society 2024
Quantum reservoir computing (QRC) provides a powerful framework for processing temporal data using quantum dynamics, but incorporating measurements into the reservoir remains a fundamental challenge and distinctive feature with respect to classical settings. The induced back-action can vary from a source of disturbance to a computational resource, as measurement deeply modifies the dynamics underlying temporal processing. Existing approaches have treated specific monitoring schemes independently, missing the common physical principles governing online quantum reservoirs. Here we develop a general theory of monitored quantum reservoir computing based on indirect quantum measurements, which unifies projective, weak, partial, and dissipative monitoring protocols within a single operational framework. Measurement back-action can serve as a controllable resource, providing the effective dissipation and non-unital dynamics required for successful QRC, even when the underlying unmonitored evolution is unsuitable. We derive general criteria under which monitored dynamics satisfy the echo-state property, fading memory, and input separability, including a necessary and sufficient condition for emergent strict contractivity. By comparing different monitoring schemes under a common reference dynamics, we show that these protocols are not interchangeable parameterizations to be optimized for peak performance, but rather constitute qualitatively distinct routes to computational capability, each enabled by the interplay between information extraction and measurement-induced disturbance -- a trade-off that can be further shaped through time multiplexing. Our results provide a unified theoretical foundation for online monitored quantum reservoir computing and establish quantum measurement engineering as a systematic approach for designing reservoir architectures across different quantum platforms.
Reservoir computing (RC) is a machine learning paradigm that harnesses dynamical systems as computational resources. In its quantum extension-quantum reservoir computing (QRC)-these principles are applied to quantum systems, whose rich dynamics broadens the landscape of information processing. In classical RC, optimal performance is typically achieved at the "edge of chaos," the boundary between order and chaos. Here, we identify its quantum many-body counterpart using the QRC implemented on the celebrated Sachdev-Ye-Kitaev model. Our analysis reveals substantial performance enhancements near two distinct characteristic "edges": a temporal boundary defined by the Thouless time, beyond which system dynamics is described by random matrix theory, and a parametric boundary governing the transition from integrable to chaotic regimes. These findings establish the "edge of many-body quantum chaos" as a design guideline for QRC.
Abstract Quantum reservoir computing (QRC) offers a powerful approach to exploit the rich dynamics of quantum systems for information processing. However, the computational performance of conventional Hermitian reservoirs is inherently constrained by their nonlinearity and information-spreading ability. In this work, we propose a non-Hermitian QRC in which non-Hermitian dynamics are employed as a tunable resource to significantly enhance the QRC performance. By incorporating an imaginary interaction term into the one-dimensional X Y spin model, the reservoir’s information propagation extends beyond the Lieb–Robinson bound, resulting in accelerated information scrambling. Through spectral analysis and memory evaluation, we demonstrate that the non-Hermitian reservoir can be tuned toward the edge of chaos by varying a single parameter that controls the non-Hermitian strength. This tuning optimizes both memory and computational capacities, which are crucial for processing temporal sequences. For applications, we evaluate the predictive performance of both classical and quantum chaotic time series. Our results demonstrate superior performance compared with the Hermitian counterpart, with particularly notable advantages in predicting signals generated by the Sachdev–Ye–Kitaev model.
Quantum reservoir computing (QRC) provides a hardware-efficient paradigm for temporal information processing on near-term quantum devices. Despite rapid experimental progress, a rigorous understanding of the structural conditions required for its scalable quantum-enhanced performance remains lacking. Here, we develop a theoretical framework in Pauli-Liouville space that provides a unified analytical treatment of the echo state property (ESP), nonlinear expressive power, and quantum resources. We first analyze the widely used qubit-resetting scheme and establish that quantum magic generated by reservoir dynamics is a necessary condition for effective computation, a requirement more fundamental than ESP. However, we prove that this architecture faces inherent expressivity limitations: all nonlinear processing originates exclusively from the classical encoding map, imposing an unavoidable trade-off between nonlinearity and memory capacity. To circumvent this structural bottleneck, we rigorously analyze Hamiltonian encoding, in which temporal inputs are embedded directly into the continuous dynamics generator. We show that the ESP is natively guaranteed by the Liouvillian spectral gap, decoupling it from quantum magic. Crucially, for any non-trivial drive Hamiltonian, the discrete-time update map exhibits a transcendental, infinite-order nonlinear dependence on the instantaneous input. Moreover, the intrinsic non-commutativity of the open-system generators governs the temporal coupling of these nonlinearities, producing highly non-separable processing of the input history. Our results establish a rigorous theoretical hierarchy of QRC architectures and provide prescriptive design principles for experiments targeting genuine quantum advantages in temporal processing.