Searcharxiv⌕ Search

arXiv subjects

Wei Luo

Publications and source records attributed to Wei Luo.

At least 127 records · Page 7Linked to original sources
arXiv2023

Adaptive Policy with Wait-$k$ Model for Simultaneous Translation

Simultaneous machine translation (SiMT) requires a robust read/write policy in conjunction with a high-quality translation model. Traditional methods rely on either a fixed wait-$k$ policy coupled with a standalone wait-$k$ translation model, or an adaptive policy jointly trained with the translation model. In this study, we propose a more flexible approach by decoupling the adaptive policy model from the translation model. Our motivation stems from the observation that a standalone multi-path wait-$k$ model performs competitively with adaptive policies utilized in state-of-the-art SiMT approaches. Specifically, we introduce DaP, a divergence-based adaptive policy, that makes read/write decisions for any translation model based on the potential divergence in translation distributions resulting from future information. DaP extends a frozen wait-$k$ model with lightweight parameters, and is both memory and computation efficient. Experimental results across various benchmarks demonstrate that our approach offers an improved trade-off between translation accuracy and latency, outperforming strong baselines.

cs.CL↗
arXiv2023

Stochastic spin-orbit-torque synapse and its application in uncertainty quantification

Stochasticity plays a significant role in the low-power operation of a biological neural network. In an artificial neural network (ANN), stochasticity also contributes to critical functions such as the uncertainty quantification (UQ) for estimating the probability for the correctness of prediction. This UQ is vital for cutting-edge applications, including medical diagnostics, autopilots, and large language models. Thanks to high computing velocity and low dissipation, a spin-orbit-torque (SOT) device exhibits significant potential for implementing the UQ. However, up until now, the application of UQ for stochastic SOT devices remains unexplored. In this study, based on SOT-induced stochastic magnetic domain wall (DW) motion with varying velocity, we fabricated an SOT synapse that could emulate stochastic weight update following the Spike-Timing-Dependent-Plasticity (STDP) rule. Furthermore, we set up a stochastic Spiking-Neural-Network (SNN), which, when compared to its deterministic counterpart, demonstrates a clear advantage in quantifying uncertainty for diagnosing the type of breast tumor (benign or malignant).

physics.app-ph↗
arXiv2023

Strongly Coupled Spin Waves and Surface Acoustic Waves at Room Temperature

Here, we report the observation of strong coupling between magnons and surface acoustic wave (SAW) phonons in a thin CoFeB film constructed in an on-chip SAW resonator by analyzing SAW phonon dispersion anticrossings. Our device design provides the tunability of the film thickness with a fixed phonon wavelength, which is a departure from the conventional approach in strong magnon--phonon coupling research. We detect a monotonic increase in the coupling strength by expanding the film thickness, which agrees with our theoretical model. Our work offers a significant way to advance fundamental research and the development of devices based on magnon--phonon hybrid quasiparticles.

cond-mat.mes-hall↗
arXiv2023

Dual Effects of the US-China Trade War and COVID-19 on United States Imports: Transfer of China's industrial chain?

The trade tension between the U.S. and China since 2018 has caused a steady decoupling of the world's two largest economies. The pandemic outbreak in 2020 complicated this process and had numerous unanticipated repercussions. This paper investigates how U.S. importers reacted to the trade war and worldwide lockdowns due to the COVID-19 pandemic. We examine the effects of the two incidents on U.S. imports separately and collectively, with various economic scopes. Our findings uncover intricate trading dynamics among the U.S., China, and Southeast Asia, through which businesses relocated portions of their global supply chain away from China to avoid high tariffs. Our analysis indicates that increased tariffs cause the U.S. to import less from China. Meanwhile, Southeast Asian exporters have integrated more into value chains centered on Chinese suppliers by participating more in assembling and completing products. However, the worldwide lockdowns over pandemic have reversed this trend as, over this period, the U.S. effectively imported more goods directly from China and indirectly through Southeast Asian exporters that imported from China.

econ.GN↗
arXiv2023

Adjusting inverse regression for predictors with clustered distribution

A major family of sufficient dimension reduction (SDR) methods, called inverse regression, commonly require the distribution of the predictor $X$ to have a linear $E(X|β^\mathsf{T}X)$ and a degenerate $\mathrm{var}(X|β^\mathsf{T}X)$ for the desired reduced predictor $β^\mathsf{T}X$. In this paper, we adjust the first and second-order inverse regression methods by modeling $E(X|β^\mathsf{T}X)$ and $\mathrm{var}(X|β^\mathsf{T}X)$ under the mixture model assumption on $X$, which allows these terms to convey more complex patterns and is most suitable when $X$ has a clustered sample distribution. The proposed SDR methods build a natural path between inverse regression and the localized SDR methods, and in particular inherit the advantages of both; that is, they are $\sqrt{n}$-consistent, efficiently implementable, directly adjustable under the high-dimensional settings, and fully recovering the desired reduced predictor. These findings are illustrated by simulation studies and a real data example at the end, which also suggest the effectiveness of the proposed methods for nonclustered data.

stat.ME↗
arXiv2023

Global strong solutions and large time behavior to the compressible co-rotation FENE dumbbell model of polymeric flows near equilibrium

In this paper, we mainly study global well-posedness and optimal decay rate for the strong solutions of the compressible co-rotation finite extensible nonlinear elastic (FENE) dumbbell model. This model is a coupling of the isentropic compressible Navier-Stokes equations with a nonlinear Fokker-Planck equation. We first prove that the FENE dumbbell model admits a unique global strong solution provided the initial data are close to equilibrium state for $d\geq 2$. Moreover, for $d\geq3$, we show that optimal decay rates of global strong solutions by the linear spectral theory and a more precise Hardy type inequality.

math.AP↗
arXiv2023

Optimal decay rate for the 2-D compressible Oldroyd-B and Hall-MHD model

In this paper, we are concerned with long time behavior of the strong solutions to the 2-D compressible Oldroyd-B and Hall-MHD model. By virtue of the improved Fourier splitting method and the time weighted energy estimate, we obtain the $L^2$ decay rate $(1+t)^{-\frac{1}{4}}$. According to the Littlewood-Paley theory, we prove that the solutions belong to the critical Besov space with negative index. Finally, we show optimal decay rate in $H^2$-framework without the smallness restriction of low frequencies.

math.AP↗
arXiv2023

Spin Swapping for an Exchange Magnon Spin Current

We propose the spin swapping effect for an exchange magnon spin current in a perpendicularly magnetized ferromagnetic medium with in plane anisotropy on the surface. The excitation of a magnon current flowing along an in-plane direction with an out-of-plane spin polarization leads to the generation of a secondary exchange spin current propagating along the out-of-plane direction, characterized by an in-plane spin polarization. The resulting exchange magnon spin current can induce an inverse spin Hall voltage of micro-volts. The exchange coupling at the interface between regions with different magnetic anisotropies plays a crucial role in generating the spin swapping effect. This is in contrast to the recently reported spin swapping for an exchange spin current in a canted antiferromagnet due to the Dzyaloshiskii-Moriya interaction.

cond-mat.mes-hall↗
arXiv2023

Optimal decay rate for the generalized Oldroyd-B model with only stress tensor diffusion in $\mathbb{R}^2$

In this paper, we are concerned with optimal decay rate for the 2-D generalized Oldroyd-B model with only stress tensor diffusion $(-Δ)^βτ$. In the case $β=1$, we first establish optimal decay rate in $H^1$ framework and remove the smallness assumption of low frequencies by virtue of the Fourier splitting method and the Littlewood-Paley decomposition theory. Furthermore, we prove optimal decay rate for the highest derivative of the solution by a different method combining time frequency decomposition and the time weighted energy estimate. In the case $\frac 1 2\leq β<1$, we study optimal decay rate for the highest derivative of the solution by the improved Fourier splitting method.

math.AP↗
arXiv2023

Do We Need an Encoder-Decoder to Model Dynamical Systems on Networks?

As deep learning gains popularity in modelling dynamical systems, we expose an underappreciated misunderstanding relevant to modelling dynamics on networks. Strongly influenced by graph neural networks, latent vertex embeddings are naturally adopted in many neural dynamical network models. However, we show that embeddings tend to induce a model that fits observations well but simultaneously has incorrect dynamical behaviours. Recognising that previous studies narrowly focus on short-term predictions during the transient phase of a flow, we propose three tests for correct long-term behaviour, and illustrate how an embedding-based dynamical model fails these tests, and analyse the causes, particularly through the lens of topological conjugacy. In doing so, we show that the difficulties can be avoided by not using embedding. We propose a simple embedding-free alternative based on parametrising two additive vector-field components. Through extensive experiments, we verify that the proposed model can reliably recover a broad class of dynamics on different network topologies from time series data.

cs.LG↗
arXiv2023

Topological linear magnetoresistivity and thermoconductivity induced by noncentrosymmetric Berry curvature

The Berry curvature plays a key role in the magnetic transport of topological materials. Yet, it is not clear whether the Berry curvature by itself can give rise to universal transport phenomena with specific scaling behaviors. In this work, based on the semiclassical Boltzmann formalism and the symmetry analysis, we show that the noncentrosymmetric distribution of the Berry curvature generally results in linear magnetoresistivity and thermoconductivity both exhibiting the B-scaling behavior. We then study such kind of topological linear magnetoresistivity in the 2D MnBi2Te4 flakes and the 3D spin-orbit-coupled electron gas, the former showing good agreement with the experimental observations. The difference between our mechanism and the conventional anisotropic magnetoresistance is elucidated. Our theory proposes a universal scenario for the topological linear magnetoresistivity and thermoconductivity and predicts such effects to occur in various materials, which also provides a reasonable explanation for the recent observations of linear magnetoresistivity.

cond-mat.mes-hall↗
arXiv2023

Visual Anomaly Detection via Dual-Attention Transformer and Discriminative Flow

In this paper, we introduce the novel state-of-the-art Dual-attention Transformer and Discriminative Flow (DADF) framework for visual anomaly detection. Based on only normal knowledge, visual anomaly detection has wide applications in industrial scenarios and has attracted significant attention. However, most existing methods fail to meet the requirements. In contrast, the proposed DTDF presents a new paradigm: it firstly leverages a pre-trained network to acquire multi-scale prior embeddings, followed by the development of a vision Transformer with dual attention mechanisms, namely self-attention and memorial-attention, to achieve two-level reconstruction for prior embeddings with the sequential and normality association. Additionally, we propose using normalizing flow to establish discriminative likelihood for the joint distribution of prior and reconstructions at each scale. The DADF achieves 98.3/98.4 of image/pixel AUROC on Mvtec AD; 83.7 of image AUROC and 67.4 of pixel sPRO on Mvtec LOCO AD benchmarks, demonstrating the effectiveness of our proposed approach.

cs.CV↗
arXiv2023

Learning Global-Local Correspondence with Semantic Bottleneck for Logical Anomaly Detection

This paper presents a novel framework, named Global-Local Correspondence Framework (GLCF), for visual anomaly detection with logical constraints. Visual anomaly detection has become an active research area in various real-world applications, such as industrial anomaly detection and medical disease diagnosis. However, most existing methods focus on identifying local structural degeneration anomalies and often fail to detect high-level functional anomalies that involve logical constraints. To address this issue, we propose a two-branch approach that consists of a local branch for detecting structural anomalies and a global branch for detecting logical anomalies. To facilitate local-global feature correspondence, we introduce a novel semantic bottleneck enabled by the visual Transformer. Moreover, we develop feature estimation networks for each branch separately to detect anomalies. Our proposed framework is validated using various benchmarks, including industrial datasets, Mvtec AD, Mvtec Loco AD, and the Retinal-OCT medical dataset. Experimental results show that our method outperforms existing methods, particularly in detecting logical anomalies.

cs.CV↗
arXiv2023

Large time behavior to a 2D micro-macro model for compressible polymeric fluids near equilibrium

In this paper, we mainly study the large time behavior to a 2D micro-macro model for compressible polymeric fluids with small initial data. This model is a coupling of isentropic compressible Navier-Stokes equations with a nonlinear Fokker-Planck equation. Firstly the Fourier splitting method yields that the logarithmic decay rate. By virtue of the time weighted energy estimate, we can improve the decay rate to $(1 + t)^{-\frac{1}{4}}$. Under the low-frequency condition and by the Littlewood-Paley theory, we show that the solutions belong to some Besov spaces with negative index and obtain the optimal $L^2$ decay rate. Finally, we obtain the $\dot{H}^s$ decay rate by establishing a new Fourier splitting estimate.

math.AP↗
arXiv2023

One Adapter for All Programming Languages? Adapter Tuning for Code Search and Summarization

As pre-trained models automate many code intelligence tasks, a widely used paradigm is to fine-tune a model on the task dataset for each programming language. A recent study reported that multilingual fine-tuning benefits a range of tasks and models. However, we find that multilingual fine-tuning leads to performance degradation on recent models UniXcoder and CodeT5. To alleviate the potentially catastrophic forgetting issue in multilingual models, we fix all pre-trained model parameters, insert the parameter-efficient structure adapter, and fine-tune it. Updating only 0.6\% of the overall parameters compared to full-model fine-tuning for each programming language, adapter tuning yields consistent improvements on code search and summarization tasks, achieving state-of-the-art results. In addition, we experimentally show its effectiveness in cross-lingual and low-resource scenarios. Multilingual fine-tuning with 200 samples per programming language approaches the results fine-tuned with the entire dataset on code summarization. Our experiments on three probing tasks show that adapter tuning significantly outperforms full-model fine-tuning and effectively overcomes catastrophic forgetting.

cs.SE↗
arXiv2023

Tailoring the plasticity of topologically close-packed phases via the crystals' fundamental building blocks

Brittle topologically close-packed precipitates form in many advanced alloys. Due to their complex structures little is known about their plasticity. Here, we present a strategy to understand and tailor the deformability of these complex phases by considering the Nb-Co μ-phase as an archetypal material. The plasticity of the Nb-Co μ-phase is controlled by the Laves phase building block that forms parts of its unit cell. We find that between the bulk C15-NbCo$_2$ Laves and Nb-Co μ-phase, the interplanar spacing and local elastic modulus of the Laves phase building block change, leading to a strong reduction in hardness and elastic modulus, as well as a transition from synchroshear to crystallographic slip. Furthermore, as the composition changes from Nb$_6$Co$_7$ to Nb$_7$Co$_6$, the Co atoms in the triple layer are substituted such that the triple layer of the Laves phase building block becomes a slab of pure Nb, resulting in inhomogeneous changes in elasticity and a transition from crystallographic slip to a glide-and-shuffle mechanism. These findings open opportunities to purposefully tailor the plasticity of these topologically close-packed phases in bulk, but at the atomic scale of interplanar spacing and local shear modulus of the fundamental crystal building blocks in their large unit cells.

cond-mat.mtrl-sci↗
arXiv2023

A New Two-Dimensional Dirac Semimetal Based on the Alkaline Earth Metal, CaP$_3$

Using an evolutionary algorithm in combination with first-principles density functional theory calculations, we identify two-dimensional (2D) CaP$_3$ monolayer as a new Dirac semimetal due to inversion and nonsymmorphic spatial symmetries of the structure. This new topological material, composed of light elements, exhibits high structural stability (higher than the phase known in the literature), which is confirmed by thermodynamic and kinetic stability analysis. Moreover, it satisfies the electron filling criteria, so that its Dirac state is located near the Fermi level. The existence of the Dirac state predicted by the theoretical symmetry analysis is also confirmed by first-principles electronic band structure calculations. We find that the energy position of the Dirac state can be tuned by strain, while the Dirac state is unstable against an external electric field since it breaks the spatial inversion symmetry. Our findings should be instrumental in the development of 2D Dirac fermions based on light elements for their application in nanoelectronic devices and topological electronics.

cond-mat.mtrl-sci↗
arXiv2023

Clear Memory-Augmented Auto-Encoder for Surface Defect Detection

In surface defect detection, due to the extreme imbalance in the number of positive and negative samples, positive-samples-based anomaly detection methods have received more and more attention. Specifically, reconstruction-based methods are the most popular. However, existing methods are either difficult to repair abnormal foregrounds or reconstruct clear backgrounds. Therefore, we propose a clear memory-augmented auto-encoder (CMA-AE). At first, we propose a novel clear memory-augmented module (CMAM), which combines the encoding and memoryencoding in a way of forgetting and inputting, thereby repairing abnormal foregrounds and preserving clear backgrounds. Secondly, a general artificial anomaly generation algorithm (GAAGA) is proposed to simulate anomalies that are as realistic and feature-rich as possible. At last, we propose a novel multi scale feature residual detection method (MSFR) for defect segmentation, which makes the defect location more accurate. Extensive comparison experiments demonstrate that CMA-AE achieves state-of-the-art detection accuracy and shows great potential in industrial applications.

cs.CV↗