arXiv subjects

Antonio Sannia

Publications and source records attributed to Antonio Sannia.

9 recordsLinked to original sources

arXiv2026

Temporal information processing on a 4,500-qubit quantum annealer

Quantum machine learning could uncover statistical structure beyond the reach of classical models, but this requires quantum models large and expressive enough to be useful and cheap enough to read out. Most approaches optimize many quantum parameters and are thus limited by expensive training loops. Here we report a quantum machine-learning model implemented on a programmable superconducting quantum annealer that processes temporal data at large scale using up to 4,500 qubits-the largest quantum machine-learning experiment performed to date. Following the quantum reservoir computing paradigm, the untrained native many-body dynamics generated by reverse annealing is directly used to process temporal data. We prove that the interactions produced during annealing are indispensable-a non-interacting reservoir retains no memory of its input. We evaluate our model experimentally on standard memory benchmarks and demonstrate that it can successfully forecast chaotic time series. These results establish quantum annealers as a scalable platform for large-scale quantum machine learning.

quant-ph↗

arXiv2026

Non-Markovianity and memory enhancement in Quantum Reservoir Computing

Featuring memory of past inputs is a fundamental requirement for machine learning models processing time-dependent data. In quantum reservoir computing, all architectures proposed so far rely on Markovian dynamics, which, as we prove, inherently lead to an exponential decay of past information, thereby limiting long-term memory capabilities. We demonstrate that non-Markovian dynamics can overcome this limitation, enabling extended memory retention. By analytically deriving memory bounds and supporting our findings with numerical simulations, we show that non-Markovian reservoirs can outperform their Markovian counterparts, particularly in tasks that require a coexistence of short- and long-term correlations. We introduce an embedding approach that allows a controlled transition from Markovian to non-Markovian evolution, providing a path for practical implementations. Our results establish quantum non-Markovianity as a key resource for enhancing memory in quantum machine learning architectures, with broad implications in quantum neural networks.

quant-ph↗

arXiv2026

A Reservoir Computing Approach to Quantum Gate Synthesis

Quantum gate synthesis is essential for implementing quantum algorithms on real hardware, yet existing methods are often computationally demanding. Here, we introduce a novel approach based on reservoir computing, which we name Group Reservoir Computing, an efficient machine-learning paradigm for learning temporal dynamics whose training reduces to a single linear regression, to reduce the resources required. The method is grounded in the Wei--Norman decomposition, which provides a compact description of the evolution. We prove that the reconstructed dynamics always remain unitary by construction and derive formal error bounds that establish the theoretical validity of the strategy. On the standard single-qubit gate set the trained network produces a control pulse in a single pass, with mean fidelity 0.94 across the eight benchmark gates; used to warm-start gradient-based optimization, it roughly halves the number of iterations that plain gradient ascent needs to reach a target fidelity, so that the relevant figure of merit is the time to reach that threshold rather than the final accuracy after a fixed budget. Owing to its general formulation, the method applies to any finite-dimensional hardware platform; the route to multiqubit synthesis is discussed in the closing section.

quant-ph↗

arXiv2026

Uncovering and Circumventing Noise in Quantum Algorithms via Metastability

The presence of noise is the primary challenge in realizing fault-tolerant quantum computers. In this work, we introduce and experimentally validate a novel strategy to circumvent noise by exploiting the phenomenon of metastability, where a dynamical system exhibits a separation of time scales in its evolution. We demonstrate that if quantum hardware noise exhibits metastability, both digital and analog algorithms can be designed in a noise-aware fashion to achieve intrinsic resilience. We develop a general theoretical framework and introduce an efficiently computable noise vulnerability metric that avoids the need for full classical simulation of the quantum algorithm. We show that the noise vulnerability index bounds errors in noisy implementations, with smaller values indicating greater fidelity between the achieved and target quantum states. We illustrate the use of our framework with applications to variational quantum algorithms and analog adiabatic state preparation. Crucially, we provide experimental evidence supporting the presence of metastable noise in gate-model quantum processors and quantum annealing devices. Thus, we establish that the noise properties in near-term quantum hardware can directly inform practical implementation strategies, enabling the preparation of final noisy states that more closely approximate the ideal ones.

quant-ph↗

arXiv2025

From quantum feature maps to quantum reservoir computing: perspectives and applications

We explore the interplay between two emerging paradigms: reservoir computing and quantum computing. We observe how quantum systems featuring beyond-classical correlations and vast computational spaces can serve as non-trivial, experimentally viable reservoirs for typical tasks in machine learning. With a focus on neutral atom quantum processing units, we describe and exemplify a novel quantum reservoir computing (QRC) workflow. We conclude exploratively discussing the main challenges ahead, whilst arguing how QRC can offer a natural candidate to push forward reservoir computing applications.

quant-ph↗

arXiv2025

Exponential concentration and symmetries in Quantum Reservoir Computing

Quantum reservoir computing (QRC) is an emerging framework for near-term quantum machine learning that offers in-memory processing, platform versatility across analogue and digital systems, and avoids typical trainability challenges such as barren plateaus and local minima. The exponential number of independent features of quantum reservoirs opens the way to a potential performance improvement compared to classical settings. However, this exponential scaling can be hindered by exponential concentration, where finite-ensemble noise in quantum measurements requires exponentially many samples to extract meaningful outputs, a common issue in quantum machine learning. In this work, we go beyond static quantum machine learning tasks and address concentration in QRC for time-series processing using quantum-scrambling reservoirs. Beyond discussing how concentration effects can constrain QRC performance, we demonstrate that leveraging Hamiltonian symmetries significantly suppresses concentration, enabling robust and scalable QRC implementations. We illustrate our approach with concrete examples, including an established QRC design.

quant-ph↗

arXiv2025

Liouvillian skin effect in quantum neural networks

In the field of dissipative systems, the non-Hermitian skin effect has generated significant interest due to its unexpected implications. A system is said to exhibit a skin effect if its properties are largely affected by the boundary conditions. Despite the burgeoning interest, the potential impact of this phenomenon on emerging quantum technologies remains unexplored. In this work, we address this gap by demonstrating that quantum neural networks can exhibit this behavior and that skin effects, beyond their fundamental interest, can also be exploited in computational tasks. Specifically, we show that the performance of a given complex network used as a quantum reservoir computer is dictated solely by the boundary conditions of a dissipative line within its architecture. The closure of one (edge) link is found to drastically change the performance in time-series processing, proving the possibility of exploiting skin effects for machine learning.

quant-ph↗

arXiv2024

Engineered dissipation to mitigate barren plateaus

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 can restore the trainability of quantum models. We identify the required form of the dissipation processes and establish that their optimization is efficient. We benchmark our proposal in both a synthetic and a practical quantum chemistry example, demonstrating its effectiveness and potential impact across different domains.

quant-ph↗

arXiv2024

Dissipation as a resource for Quantum Reservoir Computing

Dissipation induced by interactions with an external environment typically hinders the performance of quantum computation, but in some cases can be turned out as a useful resource. We show the potential enhancement induced by dissipation in the field of quantum reservoir computing introducing tunable local losses in spin network models. Our approach based on continuous dissipation is able not only to reproduce the dynamics of previous proposals of quantum reservoir computing, based on discontinuous erasing maps but also to enhance their performance. Control of the damping rates is shown to boost popular machine learning temporal tasks as the capability to linearly and non-linearly process the input history and to forecast chaotic series. Finally, we formally prove that, under non-restrictive conditions, our dissipative models form a universal class for reservoir computing. It means that considering our approach, it is possible to approximate any fading memory map with arbitrary precision.

quant-ph↗