Much attention has been devoted to the use of machine learning to approximate physical concepts. Yet, due to challenges in interpretability of machine learning techniques, the question of what physics machine learning models are able to learn remains open. Here we bridge the concept a physical quantity and its machine learning approximation in the context of the original application of neural networks in physics: topological phase classification. We construct a hybrid tensor-neural network object that exactly expresses real space topological invariant and rigorously assess its trainability and generalization. Specifically, we benchmark the accuracy and trainability of a tensor-neural network to multiple types of neural networks, thus exemplifying the differences in trainability and representational power. Our work highlights the challenges in learning topological invariants and constitutes a stepping stone towards more accurate and better generalizable machine learning representations in condensed matter physics.

The standard dynamical approach to quantum thermodynamics is based on Markovian master equations describing the thermalization of a system weakly coupled to a large environment, and on tools such as entropy production relations. Here we develop a new framework overcoming the limitations that the current dynamical and information theory approaches encounter when applied to this setting. More precisely, we introduce the notion of continuous thermomajorization, and employ it to obtain necessary and sufficient conditions for the existence of a Markovian thermal process transforming between given initial and final energy distributions of the system. These lead to a complete set of generalized entropy production inequalities including the standard one as a special case. Importantly, these conditions can be reduced to a finitely verifiable set of constraints governing non-equilibrium transformations under master equations. What is more, the framework is also constructive, i.e., it returns explicit protocols realizing any allowed transformation. These protocols use as building blocks elementary thermalizations, which we prove to be universal controls. Finally, we also present an algorithm constructing the full set of energy distributions achievable from a given initial state via Markovian thermal processes and provide a $\texttt{Mathematica}$ implementation solving $d=6$ on a laptop computer in minutes.

Neural quantum states (NQS) have emerged as a powerful tool for approximating quantum wavefunctions using deep learning. While these models achieve remarkable accuracy, understanding how they encode physical information remains an open challenge. In this work, we introduce adiabatic fine-tuning, a scheme that trains NQS across a phase diagram, leading to strongly correlated weight representations across different models. This correlation in weight space enables the detection of phase transitions in quantum systems by analyzing the trained network weights alone. We validate our approach on the transverse field Ising model and the J1-J2 Heisenberg model, demonstrating that phase transitions manifest as distinct structures in weight space. Our results establish a connection between physical phase transitions and the geometry of neural network parameters, opening new directions for the interpretability of machine learning models in physics.

A promising use of quantum networking is quantum key distribution (QKD), which can provide information-theoretic security unattainable by classical means. While optical fiber-based QKD networks suffer from exponential loss, satellite-assisted quantum communication offers a scalable solution for long-distance secure key exchange. In this work, we propose and evaluate a satellite-based QKD setup covering the Iberian Peninsula, linking Madrid with Barcelona, Bilbao, and Lisbon. Our proposed setup uses a Low-Earth-Orbit (LEO) state-of-the-art satellite equipped with a spontaneous parametric down-conversion (SPDC) source to distribute entangled photon pairs to ground stations. Considering vibrations in the satellite, we optimize the beam waist to enhance the transmission probability and improve the secret key rate (SKR). Our results show that key rates sufficient for real-world applications, such as secure communication between hospitals, using hybrid classical-quantum protocols are feasible with existing protocols. Our results highlight the viability of near-term satellite-based QKD networks for national-scale secure communications.

Understanding scattering mechanisms in semiconductor heterostructures is crucial to reducing sources of disorder and ensuring high yield and uniformity in large spin qubit arrays. Disorder of the parent two-dimensional electron or hole gas is commonly estimated by the critical, percolation-driven density associated with the metal-insulator transition. However, a reliable estimation of the critical density within percolation theory is hindered by the need to measure conductivity with high precision at low carrier densities, where experiments are most difficult. Here, we connect experimentally percolation density and quantum Hall plateau width, in line with an earlier heuristic intuition, and offer an alternative method for characterizing semiconductor heterostructure disorder.

Optimized control of quantum networks is essential for enabling distributed quantum applications with strict performance requirements. In near-term architectures with constrained hardware, effective control may determine the feasibility of deploying such applications. Because quantum network dynamics are suitable for being modeled as a Markov decision process, dynamic programming and reinforcement learning (RL) offer promising tools for optimizing control strategies. However, key quantum network performance measures -- such as secret key rate in quantum key distribution -- often involve a non-linear relationship, with interdependent variables that describe quantum state quality and generation rate. Such objectives are not easily captured by standard RL approaches based on additive rewards. We propose a novel RL framework that directly optimizes non-linear, differentiable objective functions, while also accounting for uncertainties introduced by classical communication delays. We evaluate this framework in the context of entanglement distillation between two quantum network nodes equipped with multiplexing capability, and find that, in certain parameter regimes, it discovers policies that outperform heuristic baselines. Our work comprises the first step towards non-linear objective function optimization in quantum networks with RL, opening a path towards more advanced use cases.

In an accompanying paper [1], we introduced an approach to interface trapped-ion quantum processors with ensemble-based quantum memories by matching a spontaneous parametric down conversion source to both the ions and the memories. This enables rapid entanglement generation between single trapped ions separated by distances of hundreds of kilometers. In this article, we extend the protocol and provide additional details of the analysis. Particularly, we compare a double-click and single-click approaches for the ion edge nodes. The double-click approach relaxes the phase stability requirement but is strongly affected by finite efficiencies. Choosing the optimal protocol thus depends on the access to the phase stabilization as well as the efficiency of interface of the ions and ensemble-based memories.

Reliable quantum communication over hundreds of kilometers is a daunting yet necessary requirement for a quantum internet. To overcome photon loss, the deployment of quantum repeater stations between distant network nodes is necessary. A plethora of different quantum hardware is being developed for this purpose, each platform with its own opportunities and challenges. Here, we propose to combine two promising hardware platforms in a hybrid quantum repeater architecture to lower the cost and boost the performance of long-distance quantum communication. We outline how ensemble-based quantum memories combined with single-spin photon transducers, which can transfer quantum information between a photon and a single spin, can facilitate massive multiplexing, efficient photon generation, and quantum logic for amplifying communication rates. As a specific example, we describe how a single Rubidium (Rb) atom coupled to nanophotonic resonators can function as a high-rate, telecom-visible entangled photon source with the visible photon being compatible with storage in a Thulium-doped crystal memory (Tm-memory) and the telecom photon being compatible with low loss fiber propagation. We experimentally verify that Tm and Rb transitions are in resonance with each other. Our analysis shows that by employing up to 9 repeater stations, each equipped with two Tm-memories capable of holding up to 625 storage modes, along with four single Rb atoms, one can reach a quantum communication rate of about 10 secret bits per second across distances of up to 1000 km.

Solid-state qubits are sensitive to their microscopic environment, causing the qubit properties to fluctuate on a wide range of timescales. The sub-Hz end of the spectrum is usually dealt with by repeated background calibrations, which bring considerable overhead. It is thus important to characterize and understand the low-frequency variations of the relevant qubit characteristics. In this study, we investigate the stability of spin qubit frequencies in the Si/SiGe quantum dot platform. We find that the calibrated qubit frequencies of a six-qubit device vary by up to $\pm 100$ MHz while performing a variety of experiments over a span of 912 days. These variations are sensitive to the precise voltage settings of the gate electrodes, however when these are kept constant to within 15 $\mathrm{\mu}$V, the qubit frequencies vary by less than $\pm 7$ MHz over periods up to 36 days. During overnight scans, the qubit frequencies of ten qubits across two different devices show a standard deviation below 200 kHz within a 1-hour time window. The qubit frequency noise spectral density shows roughly a $1/f$ trend above $10^{-4}$ Hz and, strikingly, a steeper trend at even lower frequencies.

Two qubit gates constitute fundamental building blocks in the realization of large-scale quantum devices. Using superconducting circuits, two-qubit gates have previously been implemented in different ways with each method aiming to maximize gate fidelity. Another important goal of a new gate scheme is to minimize the complexity of gate calibration. In this work, we demonstrate a high-fidelity two-qubit gate between two fluxonium qubits enabled by an intermediate capacitively coupled transmon. The coupling strengths between the qubits and the coupler are designed to minimize residual crosstalk while still allowing for fast gate operations. The gate is based on frequency selectively exciting the coupler using a microwave drive to complete a 2$\pi$ rotation, conditional on the state of the fluxonium qubits. When successful, this drive scheme implements a conditional phase gate. Using analytically derived pulse shapes, we minimize unwanted excitations of the coupler and obtain gate errors of $10^{-2}$ for gate times below 60~ns. At longer durations, our gate is limited by relaxation of the coupler. Our results show how carefully designed control pulses can speed up frequency selective entangling gates.

Neural-network decoders can achieve a lower logical error rate compared to conventional decoders, like minimum-weight perfect matching, when decoding the surface code. Furthermore, these decoders require no prior information about the physical error rates, making them highly adaptable. In this study, we investigate the performance of such a decoder using both simulated and experimental data obtained from a transmon-qubit processor, focusing on small-distance surface codes. We first show that the neural network typically outperforms the matching decoder due to better handling errors leading to multiple correlated syndrome defects, such as $Y$ errors. When applied to the experimental data of [Google Quantum AI, Nature 614, 676 (2023)], the neural network decoder achieves logical error rates approximately $25\%$ lower than minimum-weight perfect matching, approaching the performance of a maximum-likelihood decoder. To demonstrate the flexibility of this decoder, we incorporate the soft information available in the analog readout of transmon qubits and evaluate the performance of this decoder in simulation using a symmetric Gaussian-noise model. Considering the soft information leads to an approximately $10\%$ lower logical error rate, depending on the probability of a measurement error. The good logical performance, flexibility, and computational efficiency make neural network decoders well-suited for near-term demonstrations of quantum memories.

Design space exploration (DSE) plays an important role in optimising quantum circuit execution by systematically evaluating different configurations of compilation strategies and hardware settings. In this work, we study the impact of layout methods, qubit routing techniques, compiler optimization levels, and hardware-specific properties, including noise characteristics, topological structures, connectivity densities, and device sizes. By traversing these dimensions, we aim to understand how compilation choices interact with hardware features. A central question in our study is whether carefully selected device parameters and mapping strategies, including initial layouts and routing heuristics, can mitigate hardware-induced errors beyond standard error mitigation methods. Our results show that choosing the right software strategies (e.g., layout and routing) and tailoring hardware properties (e.g., reducing noise or leveraging connectivity) significantly enhances the fidelity of quantum circuit executions. We provide performance estimates using metrics such as circuit depth, gate count, and expected fidelity. These findings highlight the value of hardware-software co-design, especially as quantum systems scale and move toward error-corrected computing. Our simulations, though noisy, include quantum error correction (QEC) scenarios, revealing similar sensitivities to layout and connectivity. This suggests that co-design principles will be vital for integrating QEC in future devices. Overall, we offer practical guidance for co-optimizing mapping, routing, and hardware configuration in real-world quantum computing.

The goal of this tutorial is to provide an overview of the main principles behind randomized benchmarking techniques. A newcomer to the field faces the challenge that a considerable amount of background knowledge is required to get familiar with the topic. Our purpose is to ease this process by providing a pedagogical introduction to randomized benchmarking. Every chapter is supplemented with an accompanying Python notebook, illustrating the essential steps of each protocol.

We present a neural network wavefunction framework for solving non-Abelian lattice gauge theories in a continuous group representation. Using a combination of $SU(2)$ equivariant neural networks alongside an $SU(2)$ invariant, physics-inspired ansatz, we learn a parameterization of the ground state wavefunction of $SU(2)$ lattice gauge theory in 2+1 and 3+1 dimensions. Our method, performed in the Hamiltonian formulation, has a straightforward generalization to $SU(N)$. We benchmark our approach against a solely invariant ansatz by computing the ground state energy, demonstrating the need for bespoke gauge equivariant transformations. We evaluate the Creutz ratio and average Wilson loop, and obtain results in strong agreement with perturbative expansions. Our method opens up an avenue for studying lattice gauge theories beyond one dimension, with efficient scaling to larger systems, and in a way that avoids both the sign problem and any discretization of the gauge group.

Spin qubits in semiconducting quantum dots are currently limited by slow readout processes, which are orders of magnitude slower than gate operations. In contrast, Andreev spin qubits benefit from fast measurement schemes enabled by the large resonator couplings of superconducting qubits but suffer from reduced coherence during qubit operations. Here, we propose fast and high-fidelity measurement protocols based on an electrically-tunable coupling between quantum dot and Andreev spin qubits. In realistic devices, this coupling can be made sufficiently strong to enable high-fidelity readout well below microseconds, potentially enabling mid-circuit measurements. Crucially, the electrical tunability of our coupler permits to switch it off during idle periods, minimizing crosstalk and measurement back-action. Our approach is fully compatible with germanium-based devices and paves the way for scalable quantum computing architectures by leveraging the advantages of heterogeneous qubit implementations.

Protecting qubits from noise is essential for building reliable quantum computers. Topological qubits offer a route to this goal by encoding quantum information non-locally, using pairs of Majorana zero modes. These modes form a shared fermionic state whose occupation -- either even or odd -- defines the fermionic parity that encodes the qubit. Crucially, this parity cannot be accessed by any measurement that probes only one Majorana mode. This reflects the non-local nature of the encoding and its inherent protection against noise. A promising platform for realizing such qubits is the Kitaev chain, implemented in quantum dots coupled via superconductors. Even a minimal chain of two dots can host a pair of Majorana modes and store quantum information in their joint parity. Here we introduce a new technique for reading out this parity, based on quantum capacitance. This global probe senses the joint state of the chain and enables real-time, single-shot discrimination of the parity state. By comparing with simultaneous local charge sensing, we confirm that only the global signal resolves the parity. We observe random telegraph switching and extract parity lifetimes exceeding one millisecond. These results establish the essential readout step for time-domain control of Majorana qubits, resolving a long-standing experimental challenge.

Nanoscale engineered spin systems, ranging from spins on surfaces to nanographenes, provide flexible platforms to realize entangled quantum magnets from a bottom up approach. However, assessing the quantum many-body Hamiltonian realized in a specific experiment remains an exceptional open challenge, due to the difficulty of disentangling competing terms accounting for the many-body excitations. Here, we demonstrate a machine learning strategy to learn a quantum many-body spin Hamiltonian from scanning spectroscopy measurements of spin excitations. Our methodology leverages the spatially-resolved reconstruction of the many-body excitations induced by depositing quantum impurities next to the quantum magnet. We demonstrate that our algorithm allows us to predict long-range Heisenberg exchange interactions, anisotropic exchange, as well as antisymmetric Dzyaloshinskii-Moriya interaction, including in the presence of sizable noise. Our methodology establishes defect-induced spatially-resolved dynamical excitations in quantum magnets as a powerful strategy to understand the nature of quantum spin many-body models.

All-electrical baseband control of qubits facilitates scaling up quantum processors by removing issues of crosstalk and heat generation. In semiconductor quantum dots, this is enabled by multi-spin qubit encodings, such as the exchange-only qubit, where high-fidelity readout and both single- and two-qubit operations have been demonstrated. However, their performance is limited by unavoidable leakage states that are energetically close to the computational subspace. In this work, we introduce an alternative, scalable spin qubit architecture that leverages strong spin-orbit interactions of hole nanostructures for baseband qubit operations while completely eliminating leakage channels and reducing the overall gate overhead. This encoding is intrinsically robust to local variability in hole spin properties and operates with two degenerate states, removing the need for precise calibration and mitigating heat generation from fast signal sources. Finally, our architecture is fully compatible with current technology, utilizing the same initialization, readout, and multi-qubit protocols of state-of-the-art spin-1/2 systems. By addressing critical scalability challenges, our design offers a robust and scalable pathway for semiconductor spin qubit technologies.

Near-term quantum networks face a bottleneck due to low quantum communication rates. This degrades performance both by lowering operating speeds and increasing qubit storage time in noisy memories, making some quantum internet applications infeasible. One way to circumvent this bottleneck is multiplexing: combining multiple signals into a single signal to improve the overall rate. Standard multiplexing techniques are classical in that they do not make use of coherence between quantum channels nor account for decoherence rates that vary during a protocol's execution. In this paper, we first derive semiclassical limits to multiplexing for many-qubit protocols, and then introduce new techniques: quantum multiplexing and multi-server multiplexing. These can enable beyond-classical multiplexing advantages. We illustrate these techniques through three example applications: 1) entanglement generation between two asymetric quantum network nodes (i.e., repeaters or quantum servers with inequal memories), 2) remote state preparation between many end user devices and a single quantum node, and 3) remote state preparation between one end user device and many internetworked quantum nodes. By utilizing many noisy internetworked quantum devices instead of fewer low-noise devices, our multiplexing strategies enable new paths towards achieving high-speed many-qubit quantum network applications.