arXiv subjects

Xiang Chen

Publications and source records attributed to Xiang Chen.

At least 487 records · Page 27Linked to original sources

arXiv2021

Imaging antiferromagnetic domain fluctuations and the effect of atomic-scale disorder in a doped spin-orbit Mott insulator

Correlated oxides can exhibit complex magnetic patterns, characterized by domains with vastly different size, shape and magnetic moment spanning the material. Understanding how magnetic domains form in the presence of chemical disorder and their robustness to temperature variations has been of particular interest, but atomic-scale insight into this problem has been limited. We use spin-polarized scanning tunneling microscopy to image the evolution of spin-resolved modulations originating from antiferromagnetic (AF) ordering in a spin-orbit Mott insulator Sr3Ir2O7 as a function of chemical composition and temperature. We find that replacing only several percent of La for Sr leaves behind nanometer-scale AF puddles clustering away from La substitutions preferentially located in the middle SrO layer within the unit cell. Thermal erasure and re-entry into the low-temperature ground state leads to a spatial reorganization of the AF modulations, indicating multiple stable AF configurations at low temperature. Interestingly, regardless of this rearrangement, the AF puddles maintain scale-invariant fractal geometry in each configuration. Our experiments reveal spatial fluctuations of the AF order in electron doped Sr3Ir2O7, and shed light on its sensitivity to different types of atomic-scale disorder.

cond-mat.str-el↗

arXiv2021

ZJUKLAB at SemEval-2021 Task 4: Negative Augmentation with Language Model for Reading Comprehension of Abstract Meaning

This paper presents our systems for the three Subtasks of SemEval Task4: Reading Comprehension of Abstract Meaning (ReCAM). We explain the algorithms used to learn our models and the process of tuning the algorithms and selecting the best model. Inspired by the similarity of the ReCAM task and the language pre-training, we propose a simple yet effective technology, namely, negative augmentation with language model. Evaluation results demonstrate the effectiveness of our proposed approach. Our models achieve the 4th rank on both official test sets of Subtask 1 and Subtask 2 with an accuracy of 87.9% and an accuracy of 92.8%, respectively. We further conduct comprehensive model analysis and observe interesting error cases, which may promote future researches.

cs.CL↗

arXiv2021

Resource Rationing for Wireless Federated Learning: Concept, Benefits, and Challenges

We advocate a new resource allocation framework, which we term resource rationing, for wireless federated learning (FL). Unlike existing resource allocation methods for FL, resource rationing focuses on balancing resources across learning rounds so that their collective impact on the federated learning performance is explicitly captured. This new framework can be integrated seamlessly with existing resource allocation schemes to optimize the convergence of FL. In particular, a novel "later-is-better" principle is at the front and center of resource rationing, which is validated empirically in several instances of wireless FL. We also point out technical challenges and research opportunities that are worth pursuing. Resource rationing highlights the benefits of treating the emerging FL as a new class of service that has its own characteristics, and designing communication algorithms for this particular service.

eess.SP↗

arXiv2021

Effects of Pauli blocking and in-medium nucleon-nucleon cross sections on the stopping power at low-intermediate energy heavy ion collisions

Three typical algorithms of Pauli blocking in the quantum molecular dynamics type models are investigated in the nuclear matter, the nucleus and the heavy ion collisions. The calculations in nuclear matter show that the blocking ratios obtained with the three algorithms are underestimated 13-25\% compared to the analytical values of blocking ratios. For the finite nucleus, the spurious collisions occur around the surface of the nucleus owing to the defects of Pauli blocking algorithms. In the simulations of heavy ion collisions, the uncertainty of stopping power from different Pauli blocking algorithms is less than 5\%. Furthermore, the in-medium effects of nucleon-nucleon ($NN$) cross sections on the nuclear stopping power are discussed. Our results show that the transport models calculations with free $NN$ cross sections result in the stopping power decreasing with the beam energy at the beam energy less than 300 MeV/u. To increase or decrease the values of stopping power, an enhanced or suppressed model dependent in-medium $NN$ cross section is required.

nucl-th↗

arXiv2021

Structural and Magnetic Transitions in the Planar Antiferromagnet Ba$_4$Ir$_3$O$_{10}$

We report the structural and magnetic ground state properties of the monoclinic compound barium iridium oxide Ba$_4$Ir$_3$O$_{10}$ using a combination of resonant x-ray scattering, magnetometry, and thermodynamic techniques. Magnetic susceptibility exhibits a pronounced antiferromagnetic transition at $T_{\text{N}}$ $\approx$ 25K, a weaker anomaly at $T_{\text{S}}$ $\approx$ 142K, and strong magnetic anisotropy at all temperatures. Resonant elastic x-ray scattering experiments reveal a second order structural phase transition at $T_{\text{S}}$ and a magnetic transition at $T_{\text{N}}$. Both structural and magnetic superlattice peaks are observed at $L$ = half integer values. The magnetization anomaly at $T_{\text{S}}$ implies the presence of magneto-elastic coupling, which conceivably facilitates the symmetry lowering. Mean field critical scattering is observed above $T_{\text{S}}$. The magnetic structure of the antiferromagnetic ground state is discussed based on the measured magnetic superlattice peak intensity. Our study not only presents essential information for understanding the intertwined structural and magnetic properties in Ba$_4$Ir$_3$O$_{10}$, but also highlights the necessary ingredients for exploring novel ground states with octahedra trimers.

cond-mat.str-el↗

arXiv2021

Helios: Heterogeneity-Aware Federated Learning with Dynamically Balanced Collaboration

In this paper, we propose Helios, a heterogeneity-aware FL framework to tackle the straggler issue. Helios identifies individual devices' heterogeneous training capability, and therefore the expected neural network model training volumes regarding the collaborative training pace. For straggling devices, a "soft-training" method is proposed to dynamically compress the original identical training model into the expected volume through a rotating neuron training approach. With extensive algorithm analysis and optimization schemes, the stragglers can be accelerated while retaining the convergence for local training as well as federated collaboration.

cs.DC↗

arXiv2021

Automated Query Reformulation for Efficient Search based on Query Logs From Stack Overflow

As a popular Q&A site for programming, Stack Overflow is a treasure for developers. However, the amount of questions and answers on Stack Overflow make it difficult for developers to efficiently locate the information they are looking for. There are two gaps leading to poor search results: the gap between the user's intention and the textual query, and the semantic gap between the query and the post content. Therefore, developers have to constantly reformulate their queries by correcting misspelled words, adding limitations to certain programming languages or platforms, etc. As query reformulation is tedious for developers, especially for novices, we propose an automated software-specific query reformulation approach based on deep learning. With query logs provided by Stack Overflow, we construct a large-scale query reformulation corpus, including the original queries and corresponding reformulated ones. Our approach trains a Transformer model that can automatically generate candidate reformulated queries when given the user's original query. The evaluation results show that our approach outperforms five state-of-the-art baselines, and achieves a 5.6% to 33.5% boost in terms of $\mathit{ExactMatch}$ and a 4.8% to 14.4% boost in terms of $\mathit{GLEU}$.

cs.SE↗

arXiv2020

Design and Analysis of Uplink and Downlink Communications for Federated Learning

Communication has been known to be one of the primary bottlenecks of federated learning (FL), and yet existing studies have not addressed the efficient communication design, particularly in wireless FL where both uplink and downlink communications have to be considered. In this paper, we focus on the design and analysis of physical layer quantization and transmission methods for wireless FL. We answer the question of what and how to communicate between clients and the parameter server and evaluate the impact of the various quantization and transmission options of the updated model on the learning performance. We provide new convergence analysis of the well-known FedAvg under non-i.i.d. dataset distributions, partial clients participation, and finite-precision quantization in uplink and downlink communications. These analyses reveal that, in order to achieve an O(1/T) convergence rate with quantization, transmitting the weight requires increasing the quantization level at a logarithmic rate, while transmitting the weight differential can keep a constant quantization level. Comprehensive numerical evaluation on various real-world datasets reveals that the benefit of a FL-tailored uplink and downlink communication design is enormous - a carefully designed quantization and transmission achieves more than 98% of the floating-point baseline accuracy with fewer than 10% of the baseline bandwidth, for majority of the experiments on both i.i.d. and non-i.i.d. datasets. In particular, 1-bit quantization (3.1% of the floating-point baseline bandwidth) achieves 99.8% of the floating-point baseline accuracy at almost the same convergence rate on MNIST, representing the best known bandwidth-accuracy tradeoff to the best of the authors' knowledge.

cs.IT↗

arXiv2020

Cross-Domain Sentiment Classification with In-Domain Contrastive Learning

Contrastive learning (CL) has been successful as a powerful representation learning method. In this paper, we propose a contrastive learning framework for cross-domain sentiment classification. We aim to induce domain invariant optimal classifiers rather than distribution matching. To this end, we introduce in-domain contrastive learning and entropy minimization. Also, we find through ablation studies that these two techniques behaviour differently in case of large label distribution shift and conclude that the best practice is to choose one of them adaptively according to label distribution shift. The new state-of-the-art results our model achieves on standard benchmarks show the efficacy of the proposed method.

cs.CL↗

arXiv2020

Directed Graph Attention Neural Network Utilizing 3D Coordinates for Molecular Property Prediction

The prosperity of computer vision (CV) and natural language procession (NLP) in recent years has spurred the development of deep learning in many other domains. The advancement in machine learning provides us with an alternative option besides the computationally expensive density functional theories (DFT). Kernel method and graph neural networks have been widely studied as two mainstream methods for property prediction. The promising graph neural networks have achieved comparable accuracy to the DFT method for specific objects in the recent study. However, most of the graph neural networks with high precision so far require fully connected graphs with pairwise distance distribution as edge information. In this work, we shed light on the Directed Graph Attention Neural Network (DGANN), which only takes chemical bonds as edges and operates on bonds and atoms of molecules. DGANN distinguishes from previous models with those features: (1) It learns the local chemical environment encoding by graph attention mechanism on chemical bonds. Every initial edge message only flows into every message passing trajectory once. (2) The transformer blocks aggregate the global molecular representation from the local atomic encoding. (3) The position vectors and coordinates are used as inputs instead of distances. Our model has matched or outperformed most baseline graph neural networks on QM9 datasets even without thorough hyper-parameters searching. Moreover, this work suggests that models directly utilizing 3D coordinates can still reach high accuracies for molecule representation even without rotational and translational invariance incorporated.

cs.LG↗

arXiv2020

Towards Latency-aware DNN Optimization with GPU Runtime Analysis and Tail Effect Elimination

Despite the superb performance of State-Of-The-Art (SOTA) DNNs, the increasing computational cost makes them very challenging to meet real-time latency and accuracy requirements. Although DNN runtime latency is dictated by model property (e.g., architecture, operations), hardware property (e.g., utilization, throughput), and more importantly, the effective mapping between these two, many existing approaches focus only on optimizing model property such as FLOPS reduction and overlook the mismatch between DNN model and hardware properties. In this work, we show that the mismatch between the varied DNN computation workloads and GPU capacity can cause the idle GPU tail effect, leading to GPU under-utilization and low throughput. As a result, the FLOPs reduction cannot bring effective latency reduction, which causes sub-optimal accuracy versus latency trade-offs. Motivated by this, we propose a GPU runtime-aware DNN optimization methodology to eliminate such GPU tail effect adaptively on GPU platforms. Our methodology can be applied on top of existing SOTA DNN optimization approaches to achieve better latency and accuracy trade-offs. Experiments show 11%-27% latency reduction and 2.5%-4.0% accuracy improvement over several SOTA DNN pruning and NAS methods, respectively

cs.AR↗

arXiv2020

Short-range nematic fluctuations in Sr1-xNaxFe2As2 superconductors

Interactions between nematic fluctuations, magnetic order and superconductivity are central to the physics of iron-based superconductors. Here we report on in-plane transverse acoustic phonons in hole-doped Sr$_{1-x}$Na$_x$Fe$_2$As$_2$ measured via inelastic X-ray scattering, and extract both the nematic susceptibility and the nematic correlation length. By a self-contained method of analysis, for the underdoped ($x=0.36$) sample, which harbors a magnetically-ordered tetragonal phase, we find it hosts a short nematic correlation length $ξ$ ~ 10 $Å$ and a large nematic susceptibility $χ_{\rm nem}$. The optimal-doped ($x=0.55$) sample exhibits weaker phonon softening effects, indicative of both reduced $ξ$ and $χ_{\rm nem}$. Our results suggest short-range nematic fluctuations may favor superconductivity, placing emphasis on the nematic correlation length for understanding the iron-based superconductors.

cond-mat.supr-con↗

arXiv2020

Adversarial Evaluation of Autonomous Vehicles in Lane-Change Scenarios

Autonomous vehicles must be comprehensively evaluated before deployed in cities and highways. However, most existing evaluation approaches for autonomous vehicles are static and lack adaptability, so they are usually inefficient in generating challenging scenarios for tested vehicles. In this paper, we propose an adaptive evaluation framework to efficiently evaluate autonomous vehicles in adversarial environments generated by deep reinforcement learning. Considering the multimodal nature of dangerous scenarios, we use ensemble models to represent different local optimums for diversity. We then utilize a nonparametric Bayesian method to cluster the adversarial policies. The proposed method is validated in a typical lane-change scenario that involves frequent interactions between the ego vehicle and the surrounding vehicles. Results show that the adversarial scenarios generated by our method significantly degrade the performance of the tested vehicles. We also illustrate different patterns of generated adversarial environments, which can be used to infer the weaknesses of the tested vehicles.

cs.RO↗

arXiv2020

Third ArchEdge Workshop: Exploring the Design Space of Efficient Deep Neural Networks

This paper gives an overview of our ongoing work on the design space exploration of efficient deep neural networks (DNNs). Specifically, we cover two aspects: (1) static architecture design efficiency and (2) dynamic model execution efficiency. For static architecture design, different from existing end-to-end hardware modeling assumptions, we conduct full-stack profiling at the GPU core level to identify better accuracy-latency trade-offs for DNN designs. For dynamic model execution, different from prior work that tackles model redundancy at the DNN-channels level, we explore a new dimension of DNN feature map redundancy to be dynamically traversed at runtime. Last, we highlight several open questions that are poised to draw research attention in the next few years.

cs.AR↗

arXiv2020

Efficient Neural Network Implementation with Quadratic Neuron

Previous works proved that the combination of the linear neuron network with nonlinear activation functions (e.g. ReLu) can achieve nonlinear function approximation. However, simply widening or deepening the network structure will introduce some training problems. In this work, we are aiming to build a comprehensive second-order CNN implementation framework that includes neuron/network design and system deployment optimization.

cs.NI↗

arXiv2020

Learning color space adaptation from synthetic to real images of cirrus clouds

Cloud segmentation plays a crucial role in image analysis for climate modeling. Manually labeling the training data for cloud segmentation is time-consuming and error-prone. We explore to train segmentation networks with synthetic data due to the natural acquisition of pixel-level labels. Nevertheless, the domain gap between synthetic and real images significantly degrades the performance of the trained model. We propose a color space adaptation method to bridge the gap, by training a color-sensitive generator and discriminator to adapt synthetic data to real images in color space. Instead of transforming images by general convolutional kernels, we adopt a set of closed-form operations to make color-space adjustments while preserving the labels. We also construct a synthetic-to-real cirrus cloud dataset SynCloud and demonstrate the adaptation efficacy on the semantic segmentation task of cirrus clouds. With our adapted synthetic data for training the semantic segmentation, we achieve an improvement of 6:59% when applied to real images, superior to alternative methods.

cs.CV↗

arXiv2020

Cross-Domain Sentiment Classification with Contrastive Learning and Mutual Information Maximization

Contrastive learning (CL) has been successful as a powerful representation learning method. In this work we propose CLIM: Contrastive Learning with mutual Information Maximization, to explore the potential of CL on cross-domain sentiment classification. To the best of our knowledge, CLIM is the first to adopt contrastive learning for natural language processing (NLP) tasks across domains. Due to scarcity of labels on the target domain, we introduce mutual information maximization (MIM) apart from CL to exploit the features that best support the final prediction. Furthermore, MIM is able to maintain a relatively balanced distribution of the model's prediction, and enlarges the margin between classes on the target domain. The larger margin increases our model's robustness and enables the same classifier to be optimal across domains. Consequently, we achieve new state-of-the-art results on the Amazon-review dataset as well as the airlines dataset, showing the efficacy of our proposed method CLIM.

cs.CL↗

arXiv2020

Fourth-Order Nonlocal Tensor Decomposition Model for Spectral Computed Tomography

Spectral computed tomography (CT) can reconstruct spectral images from different energy bins using photon counting detectors (PCDs). However, due to the limited photons and counting rate in the corresponding spectral fraction, the reconstructed spectral images usually suffer from severe noise. In this paper, a fourth-order nonlocal tensor decomposition model for spectral CT image reconstruction (FONT-SIR) method is proposed. Similar patches are collected in both spatial and spectral dimensions simultaneously to form the basic tensor unit. Additionally, principal component analysis (PCA) is applied to extract latent features from the patches for a robust and efficient similarity measure. Then, low-rank and sparsity decomposition is performed on the produced fourth-order tensor unit, and the weighted nuclear norm and total variation (TV) norm are used to enforce the low-rank and sparsity constraints, respectively. The alternating direction method of multipliers (ADMM) is adopted to optimize the objective function. The experimental results with our proposed FONT-SIR demonstrates a superior qualitative and quantitative performance for both simulated and real data sets relative to several state-of-the-art methods, in terms of noise suppression and detail preservation.

physics.med-ph↗