Search

Variational quantum algorithms is one of the most representative algorithms in quantum computing, which has a wide range of applications in quantum machine learning, quantum simulation and other related fields. However, they face challenges associated with the barren plateau phenomenon, especially when dealing with large numbers of qubits, deep circuit layers, or global cost functions, making them often untrainable. In this paper, we propose a novel parameter initialization strategy based on Gaussian Mixture Models. We rigorously prove that, the proposed initialization method consistently avoids the barren plateaus problem for hardware-efficient ansatz with arbitrary length and qubits and any given cost function. Specifically, we find that the gradient norm lower bound provided by the proposed method is independent of the number of qubits $N$ and increases with the circuit depth $L$. Our results strictly highlight the significance of Gaussian Mixture model initialization strategies in determining the trainability of quantum circuits, which provides valuable guidance for future theoretical investigations and practical applications.

In this paper, we present LiGNN, a deployed large-scale Graph Neural Networks (GNNs) Framework. We share our insight on developing and deployment of GNNs at large scale at LinkedIn. We present a set of algorithmic improvements to the quality of GNN representation learning including temporal graph architectures with long term losses, effective cold start solutions via graph densification, ID embeddings and multi-hop neighbor sampling. We explain how we built and sped up by 7x our large-scale training on LinkedIn graphs with adaptive sampling of neighbors, grouping and slicing of training data batches, specialized shared-memory queue and local gradient optimization. We summarize our deployment lessons and learnings gathered from A/B test experiments. The techniques presented in this work have contributed to an approximate relative improvements of 1% of Job application hearing back rate, 2% Ads CTR lift, 0.5% of Feed engaged daily active users, 0.2% session lift and 0.1% weekly active user lift from people recommendation. We believe that this work can provide practical solutions and insights for engineers who are interested in applying Graph neural networks at large scale.

Nuclear spin memories of divalent neutral atoms can allow spin-preserving resolved-sideband cooling in a strong magnetic field [Phys. Rev. Lett. 99, 123001 (2007)]. We present a theory for cooling $^{87}$Sr nuclear-spin qubits in a weak magnetic field. The theory depends on laser excitation of $5s5p~^1P_1$ to a nearby state which results in $m_J$-dependent AC Stark shifts large compared to the hyperfine interaction. This effectively suppresses the nuclear-spin mixing due to the hyperfine interaction. Sideband cooling via the clock state quenched by the AC Stark-shifted $^1P_1$ state leads to nuclear-spin-preserving spontaneous emission back to the ground state. More than being compatible with low magnetic fields, the theory is applicable when the nuclear spin qubits are defined by the two lowest Zeeman substates., Comment: 16 pages, 2 figures

Nuclear-spin entangling gates with divalent atoms can be executed by one global laser pulse when $\Delta_{\text{Z}}<\Omega$, where $\Delta_{\text{Z}}$ is the Zeeman-splitting-dominated frequency difference for the clock-Rydberg transitions of the two nuclear-spin qubit states and $\Omega$ is the maximal Rabi frequency. Concerning the sensitivity of Rydberg-state energy to magnetic fluctuation, the gate is compatible with large-scale atomic arrays for weaker magnetic field is suitable for ensuring uniform field in a large qubit array. The gate can have a high fidelity because the relaxation and dephasing of Rydberg states, which limit the fidelity and grow with $1/\Omega$, can be mitigated with easily attainable large $\Omega$., Comment: 7 pages, 4 figures

Hydride precipitation in zirconium cladding materials can damage their integrity and durability.Service temperature and material defects have a significant effect on the dynamic growth of hydrides. In this study, we have developed a phase field model based on the assumption of elastic behaviour within a specific temperature range (613-653K). This model allows us to study the influence of temperature and interfacial effects on the morphology, stress, and average growth rate of zirconium hydride. The results suggest that changes in temperature and interfacial energy influence the aspect ratio and average growth rate of the hydride morphology. The ultimate determinant of hydride orientation is the loss of interfacial coherence, primarily induced by interfacial dislocation defects and quantifiable by the mismatch degree $q$. An escalation in interfacial coherence loss leads to a transition of hydride growth from horizontal to vertical, accompanied by the onset of redirection behaviour. Interestingly, redirection occurs at a critical mismatch level, denoted $q_c$, and remains unaffected by variations in temperature and interfacial energy. However, this redirection leads to an increase in the maximum stress, which may influence the direction of hydride crack propagation. This research highlights the importance of interfacial coherence and provides valuable insights into the morphology and growth kinetics of hydrides in zirconium alloys., Comment: 13 pages (text), 8 figures (text), 1 table (text), 6 pages (SI), 2 figures (SI)

Sparse reward environments are known to be challenging for reinforcement learning agents. In such environments, efficient and scalable exploration is crucial. Exploration is a means by which an agent gains information about the environment. We expand on this topic and propose a new intrinsic reward that systemically quantifies exploratory behavior and promotes state coverage by maximizing the information content of a trajectory taken by an agent. We compare our method to alternative exploration based intrinsic reward techniques, namely Curiosity Driven Learning and Random Network Distillation. We show that our information theoretic reward induces efficient exploration and outperforms in various games, including Montezuma Revenge, a known difficult task for reinforcement learning. Finally, we propose an extension that maximizes information content in a discretely compressed latent space which boosts sample efficiency and generalizes to continuous state spaces., Comment: 12 pages, 12 figures

We demonstrate equivalence between the reinforcement learning problem and the supervised classification problem. We consequently equate the exploration exploitation trade-off in reinforcement learning to the dataset imbalance problem in supervised classification, and find similarities in how they are addressed. From our analysis of the aforementioned problems we derive a novel loss function for reinforcement learning and supervised classification. Scope Loss, our new loss function, adjusts gradients to prevent performance losses from over-exploitation and dataset imbalances, without the need for any tuning. We test Scope Loss against SOTA loss functions over a basket of benchmark reinforcement learning tasks and a skewed classification dataset, and show that Scope Loss outperforms other loss functions., Comment: 11 pages, 2 figures, under review for NeurIPS 2023

Single domain generalization aims to enhance the ability of the model to generalize to unknown domains when trained on a single source domain. However, the limited diversity in the training data hampers the learning of domain-invariant features, resulting in compromised generalization performance. To address this, data perturbation (augmentation) has emerged as a crucial method to increase data diversity. Nevertheless, existing perturbation methods often focus on either image-level or feature-level perturbations independently, neglecting their synergistic effects. To overcome these limitations, we propose CPerb, a simple yet effective cross-perturbation method. Specifically, CPerb utilizes both horizontal and vertical operations. Horizontally, it applies image-level and feature-level perturbations to enhance the diversity of the training data, mitigating the issue of limited diversity in single-source domains. Vertically, it introduces multi-route perturbation to learn domain-invariant features from different perspectives of samples with the same semantic category, thereby enhancing the generalization capability of the model. Additionally, we propose MixPatch, a novel feature-level perturbation method that exploits local image style information to further diversify the training data. Extensive experiments on various benchmark datasets validate the effectiveness of our method., Comment: Accepted by IEEE TCSVT

Hydrogen-induced multi-spot corrosion on the surface of polycrystalline rare metals is a complex process, which involves the interactions between phases (metal, hydride and oxide), grain orientations, grain boundaries, and corrosion spots. To accurately simulate this process and comprehend the underlying physics, a theoretical method is required that includes the following mechanisms: i) hydrogen diffusion, ii) phase transformation, iii) elastic interactions between phases, especially, the interactions between the oxide film and the hydride, iv) elastic interactions between grains, and v) interactions between hydrogen solutes and grain boundaries. In this study, we report a multiphase-field model that incorporates all these requirements, and conduct a comprehensive study of hydrogen-induced spot corrosion on the uranium metal surface, including the investigation of the oxide film, multi-spot corrosion, grain orientation, and grain boundary in the monocrystal, bicrystal, and polycrystal systems. The results indicate that the oxide film can inhibit the growth of hydrides and plays a crucial role in determining the correct morphology of the hydride at the triple junction of phases. The elastic interaction between multiple corrosion spots causes the merging of corrosion spots and promotes the growth of hydrides. The introduction of grain orientations and grain boundaries results in a variety of intriguing intracrystalline and intergranular hydride morphologies. The model presented here is generally applicable to the hydrogen-induced multi-spot corrosion on any rare metal surface., Comment: 22 pages (text), 16 figures (text), 2 tables (text), 9 pages (SI), 13 figures (SI)

A sensitivity study for the search for the charged lepton flavor violating process $\tau \to \gamma\mu$ at the Super $\tau$-Charm Facility is performed with a fast simulation. With the expected performance of the current detector design and an integrated luminosity of \SI{1}{ab^{-1}} corresponding to one-year of data taking, the sensitivity on the branching fraction (BF) of $\tau \to \gamma\mu$ is estimated to be at the level of \num{e-8}. The sensitivity under different detector performances are also studied. With ideal performance, the BF could be probed to be \num{2.8e-8} at \SI{90}{\percent} confidence level. The sensitivity is expected to scale with the square root of the luminosity, therefore with a total luminosity of \SI{10}{ab^{-1}} corresponding to ten-year of data taking, the sensitivity could reach \num{8.8e-9}, which is about one order of magnitude improvement upon the current best upper limit.

Tensor networks, which have been traditionally used to simulate many-body physics, have recently gained significant attention in the field of machine learning due to their powerful representation capabilities. In this work, we propose a density-based clustering algorithm inspired by tensor networks. We encode classical data into tensor network states on an extended Hilbert space and train the tensor network states to capture the features of the clusters. Here, we define density and related concepts in terms of fidelity, rather than using a classical distance measure. We evaluate the performance of our algorithm on six synthetic data sets, four real world data sets, and three commonly used computer vision data sets. The results demonstrate that our method provides state-of-the-art performance on several synthetic data sets and real world data sets, even when the number of clusters is unknown. Additionally, our algorithm performs competitively with state-of-the-art algorithms on the MNIST, USPS, and Fashion-MNIST image data sets. These findings reveal the great potential of tensor networks for machine learning applications.

Quantum gates and entanglement based on dipole-dipole interactions of neutral Rydberg atoms are relevant to both fundamental physics and quantum information science. The precision and robustness of the Rydberg-mediated entanglement protocols are the key factors limiting their applicability in experiments and near-future industry. There are various methods for generating entangling gates by exploring the Rydberg interactions of neutral atoms, each equipped with its own strengths and weaknesses. The basics and tricks in these protocols are reviewed, with specific attention paid to the achievable fidelity and the robustness to the technical issues and detrimental innate factors., Comment: 57 pages, 10 figures

Entanglement between atoms and microwave photons in a superconducting coplanar waveguide~(SCW) can enable hybrid quantum devices and interface static and flying qubits. We study a one-step controlled-Z~(C$_{\text{Z}}$) gate between a neutral atom trapped near a SCW and a microwave mode in the SCW, which is an extension of the gate proposed in [J. D. Pritchard, et.al., Phys. Rev. A 89, 010301(R) (2014)]. The gate protocol is simple and requires one laser pulse for exciting a transition between the ground and Rydberg states of the neutral atom., Comment: 7 pages, 6 figures

Hyperentanglement~(HE), the simultaneous entanglement between two particles in more than one degrees of freedom, is relevant to both fundamental physics and quantum technology. Previous study on HE has been focusing on photons. Here, we study HE in individual neutral atoms. In most alkaline-earth-like atoms with two valence electrons and a nonzero nuclear spin, there are two stable electronic states, the ground state and the long-lived clock state, which can define an electronic qubit. Meanwhile, their nuclear spin states can define a nuclear qubit. By the Rydberg blockade effect, we show that the controlled-Z~(C$_{\text{Z}}$) operation can be generated in the electronic qubits of two nearby atoms, and simultaneously in their nuclear qubits as well, leading to a C$_{\text{Z}}\otimes$C$_{\text{Z}}$ operation which is capable to induce HE. The possibility to induce HE in individual neutral atoms offers new opportunities to study quantum science and technology based on neutral atoms., Comment: 19 pages, 10 figures

Neutral atoms are promising for large-scale quantum computing, but accurate neutral-atom entanglement depends on large Rydberg interactions which strongly limit the interatomic distances. Via a phase accumulation in detuned Rabi cycles enabled by a Rydberg interaction of similar magnitude to the Rydberg Rabi frequency, we study a controlled-phase gate with an arbitrary phase and extend it to the controlled-NOT gate. The gates need only three steps for coupling one Rydberg state, depend on easily accessible van der Waals interaction that naturally arises between distant atoms, and have no rotation error in the weak interaction regime. Importantly, they can work with very weak interactions so that well-separated qubits can be entangled. The gates are sensitive to the irremovable fluctuation of Rydberg interactions, but can still have a fidelity over 98\% with realistic position fluctuation of qubits separated over 20~$\mu$m., Comment: 7 pages, 3 figures

We present fast Rydberg-mediated entanglement involving nuclear spins of divalent atoms with $^{171}$Yb as an example. First, we show a nuclear-spin controlled phase gate of an arbitrary phase realizable either with two laser pulses when assisted by Stark shifts, or with three pulses. Second, we propose to create a state $(\lvert\text{cc}\rangle_{\text{e}} \otimes \lvert\Phi\rangle_{\text{n}} + \lvert\Phi\rangle_{\text{e}} \otimes \lvert\Psi\rangle_{\text{n}} )/\sqrt{2}$ entangled between the electrons~(e) and nuclear spins~(n) of two atoms, where $\lvert\Phi\rangle$ and $\lvert\Psi\rangle$ are two orthogonal Bell states and $\lvert \text{c}\rangle_{\text{e}}$ denotes the clock state. For want of a better term, it is called a Super Bell State for it mimics a ``large'' Bell state incorporating three ``smaller'' Bell states. Third, we show a protocol to create a three-atom state $(\sqrt{3}\lvert\text{ccc}\rangle_{\text{e}} \otimes \lvert\Lambda\rangle_{\text{n}} + \lvert \text{W}\rangle_{\text{e}} \otimes \lvert \text{GHZ}\rangle_{\text{n}} )/2$, where $\lvert\Lambda\rangle_{\text{n}}$ is a nuclear-spin state, $\lvert \text{W}\rangle_{\text{e}}$ is a W state in the ground-clock state space, and $\lvert \text{GHZ}\rangle_{\text{n}}$ is the Greenberger-Horne-Zeilinger~(GHZ) state in the nuclear-spin state space. The four protocols possess high intrinsic fidelities, do not require single-site Rydberg addressing, and can be executed with large $\Omega_{\text{m}}$ in a weak, Gauss-scale magnetic field for they involve Rydberg excitation of both nuclear-spin qubit states in each atom. The latter two protocols can enable measurement-based preparation of Bell, hyperentangled, and GHZ states., Comment: 19 pages, 11 figures