Searcharxiv⌕ Search

arXiv subjects

Wei Wu

Publications and source records attributed to Wei Wu.

At least 559 records · Page 31Linked to original sources
arXiv2016

Renormalization of trace distance and multipartite entanglement close to the quantum phase transitions of one- and two-dimensional spin-chain systems

We investigate the quantum phase transitions of spin systems in one and two dimensions by employing trace distance and multipartite entanglement along with real-space quantum renormalization group method. As illustration examples, a one-dimensional and a two-dimensional $XY$ models are considered. It is shown that the quantum phase transitions of these spin-chain systems can be revealed by the singular behaviors of the first derivatives of renormalized trace distance and multipartite entanglement in the thermodynamics limit. Moreover, we find the renormalized trace distance and multipartite entanglement obey certain universal exponential-type scaling laws in the vicinity of the quantum critical points.

quant-ph↗
arXiv2016

Response Selection with Topic Clues for Retrieval-based Chatbots

We consider incorporating topic information into message-response matching to boost responses with rich content in retrieval-based chatbots. To this end, we propose a topic-aware convolutional neural tensor network (TACNTN). In TACNTN, matching between a message and a response is not only conducted between a message vector and a response vector generated by convolutional neural networks, but also leverages extra topic information encoded in two topic vectors. The two topic vectors are linear combinations of topic words of the message and the response respectively, where the topic words are obtained from a pre-trained LDA model and their weights are determined by themselves as well as the message vector and the response vector. The message vector, the response vector, and the two topic vectors are fed to neural tensors to calculate a matching score. Empirical study on a public data set and a human annotated data set shows that TACNTN can significantly outperform state-of-the-art methods for message-response matching.

cs.CL↗
arXiv2016

Topic Aware Neural Response Generation

We consider incorporating topic information into the sequence-to-sequence framework to generate informative and interesting responses for chatbots. To this end, we propose a topic aware sequence-to-sequence (TA-Seq2Seq) model. The model utilizes topics to simulate prior knowledge of human that guides them to form informative and interesting responses in conversation, and leverages the topic information in generation by a joint attention mechanism and a biased generation probability. The joint attention mechanism summarizes the hidden vectors of an input message as context vectors by message attention, synthesizes topic vectors by topic attention from the topic words of the message obtained from a pre-trained LDA model, and let these vectors jointly affect the generation of words in decoding. To increase the possibility of topic words appearing in responses, the model modifies the generation probability of topic words by adding an extra probability item to bias the overall distribution. Empirical study on both automatic evaluation metrics and human annotations shows that TA-Seq2Seq can generate more informative and interesting responses, and significantly outperform the-state-of-the-art response generation models.

cs.CL↗
arXiv2016

Controlling Feynman diagrammatic expansions: physical nature of the pseudo gap in the two-dimensional Hubbard model

We introduce a method for summing Feynman's perturbation series based on diagrammatic Monte Carlo that significantly improves its convergence properties. This allows us to investigate in a controllable manner the pseudogap regime of the Hubbard model and to study the nodal/antinodal dichotomy at low doping and intermediate coupling. Marked differences from the weak coupling scenario are manifest, such as a higher degree of incoherence at the antinodes than at the `hot spots'. Our results show that the pseudogap and reduction of quasiparticle coherence at the antinode is due to antiferromagnetic spin correlations centered around the commensurate $(π,π)$ wavevector. In contrast, the dominant source of scattering at the node is associated with incommensurate momentum transfer. Umklapp scattering is found to play a key role in the nodal/antinodal dichotomy.

cond-mat.str-el↗
arXiv2016

Transversal fluctuations for a first passage percolation model

We introduce a new first passage percolation model in a Poissonian environment on $\mathbb{R}^{2}$. In this model, the action of a path depends on the geometry of the path and the travel time. We prove that the transversal fluctuation exponent for point-to-line action minimizers is at least $3/5$.

math.PR↗
arXiv2016

Gaussian fluctuations for the classical XY model

We study the classical XY model in bounded domains of $\mathbb{Z}^{d}$ with Dirichlet boundary conditions. We prove that when the temperature goes to zero faster than a certain rate as the lattice spacing goes to zero, the fluctuation field converges to a standard Gaussian white noise. This and related results also apply to a large class of gradient field models.

math.PR↗
arXiv2016

Tailoring reflection of graphene plasmons by focused ion beams

Graphene plasmons are of remarkable features that make graphene plasmon elements promising for applications to integrated photonic devices. The fabrication of graphene plasmon components and control over plasmon propagating are of fundamental important. Through near-field plasmon imaging, we demonstrate controllable modifying of the reflection of graphene plasmon at boundaries etched by ion beams. Moreover, by varying ion dose at a proper value, nature like reflection boundary can be obtained. We also investigate the influence of ion beam incident angle on plasmon reflection. To illustrate the application of ion beam etching, a simple graphene wedge-shape plasmon structure is fabricated and performs excellently, proving this technology as a simple and efficient tool for controlling graphene plasmons.

physics.optics↗
arXiv2016

A Novel Biologically Mechanism-Based Visual Cognition Model--Automatic Extraction of Semantics, Formation of Integrated Concepts and Re-selection Features for Ambiguity

Integration between biology and information science benefits both fields. Many related models have been proposed, such as computational visual cognition models, computational motor control models, integrations of both and so on. In general, the robustness and precision of recognition is one of the key problems for object recognition models. In this paper, inspired by features of human recognition process and their biological mechanisms, a new integrated and dynamic framework is proposed to mimic the semantic extraction, concept formation and feature re-selection in human visual processing. The main contributions of the proposed model are as follows: (1) Semantic feature extraction: Local semantic features are learnt from episodic features that are extracted from raw images through a deep neural network; (2) Integrated concept formation: Concepts are formed with local semantic information and structural information learnt through network. (3) Feature re-selection: When ambiguity is detected during recognition process, distinctive features according to the difference between ambiguous candidates are re-selected for recognition. Experimental results on hand-written digits and facial shape dataset show that, compared with other methods, the new proposed model exhibits higher robustness and precision for visual recognition, especially in the condition when input samples are smantic ambiguous. Meanwhile, the introduced biological mechanisms further strengthen the interaction between neuroscience and information science.

cs.CV↗
arXiv2016

Biologically inspired model simulating visual pathways and cerebellum function in human - Achieving visuomotor coordination and high precision movement with learning ability

In recent years, the interdisciplinary research between information science and neuroscience has been a hotspot. In this paper, based on recent biological findings, we proposed a new model to mimic visual information processing, motor planning and control in central and peripheral nervous systems of human. Main steps of the model are as follows: 1) Simulating "where" pathway in human: the Selective Search method is applied to simulate the function of human dorsal visual pathway to localize object candidates; 2) Simulating "what" pathway in human: a Convolutional Deep Belief Network is applied to simulate the hierarchical structure and function of human ventral visual pathway for object recognition; 3) Simulating motor planning process in human: habitual motion planning process in human is simulated, and motor commands are generated from the combination of control signals from past experiences; 4) Simulating precise movement control in human: calibrated control signals, which mimic the adjustment for movement from cerebellum in human, are generated and updated from calibration of movement errors in past experiences, and sent to the movement model to achieve high precision. The proposed framework mimics structures and functions of human recognition, visuomotor coordination and precise motor control. Experiments on object localization, recognition and movement control demonstrate that the new proposed model can not only accomplish visuomotor coordination tasks, but also achieve high precision movement with learning ability. Meanwhile, the results also prove the validity of the introduced mechanisms. Furthermore, the proposed model could be generalized and applied to other systems, such as mechanical and electrical systems in robotics, to achieve fast response, high precision movement with learning ability.

cs.RO↗
arXiv2016

Theory of plasmonic metasurfaces

In this paper we derive an impedance boundary condition to approximate the optical scattering effect of an array of plasmonic nanoparticles mounted on a perfectly conducting plate. We show that at some resonant frequencies the impedance blows up, allowing for a significant reduction of the scattering from the plate. Using the spectral properties of a Neumann-Poincare type operator, we investigate the dependency of the impedance with respect to changes in the nanoparticle geometry and configuration.

math.AP↗
arXiv2016

Large Deviation Principle For Finite-State Mean Field Interacting Particle Systems

We establish a large deviation principle for the empirical measure process associated with a general class of finite-state mean field interacting particle systems with Lipschitz continuous transition rates that satisfy a certain ergodicity condition. The approach is based on a variational representation for functionals of a Poisson random measure. Under an appropriate strengthening of the ergodicity condition, we also prove a locally uniform large deviation principle. The main novelty is that more than one particle is allowed to change its state simultaneously, and so a standard approach to the proof based on a change of measure with respect to a system of independent particles is not possible. The result is shown to be applicable to a wide range of models arising from statistical physics, queueing systems and communication networks. Along the way, we establish a large deviation principle for a class of jump Markov processes on the simplex, whose rates decay to zero as they approach the boundary of the domain. This result may be of independent interest.

math.PR↗
arXiv2016

Energy Harvesting in Secure MIMO Systems

The problems of energy harvesting in wireless com- munication systems have recently drawn much attention. In this paper, we focus on the investigation of energy harvesting maximization (EHM) in the important secrecy multi-input multi- output (MIMO) systems where little research has been done due to their complexity. Particularly, this paper studies the resource allocation strategies in MIMO wiretap channels, wherein we attempt to maximize the harvested energy by one or multiple multi-antenna energy receivers (ERs) (potential eavesdropper) while guaranteeing the secure communication for the multi- antenna information receiver (IR). Two types of IR, with and without the capability to cancel the interference from energy signals, are taken into account. In the scenario of single energy receiver (ER), we consider the joint design of the transmis- sion information and energy covariances for EHM. Both of the optimization problems for the two types of IR are non- convex, and appear to be difficult. To circumvent them, the combination of first order Taylor approximation and sequential convex optimization approach is proposed. Then, we extend our attention to the scenario with multiple ERs, where the artificial noise (AN) aided weighted sum-energy harvesting maximization (WS-EHM) problem is considered. Other than the approaches adopted in solving the EHM problems, an algorithm conducts in an alternating fashion is proposed to handle this problem. In particular, we first perform a judicious transformation of the WS- EHM problem. Then, a block Gauss-Seidel (GS) algorithm based on logarithmic barrier method and gradient projection (GP) is derived to obtain the optimal solution of the reformulation by solving convex problems alternately. Furthermore, the resulting block GS method is proven to converge to a Karush-Kuhn-Tucker (KKT) point of the original WS-EHM problem...

cs.IT↗
arXiv2015

Critical Percolation and the Minimal Spanning Tree in Slabs

The minimal spanning forest on $\mathbb{Z}^{d}$ is known to consist of a single tree for $d \leq 2$ and is conjectured to consist of infinitely many trees for large $d$. In this paper, we prove that there is a single tree for quasi-planar graphs such as $\mathbb{Z}^{2}\times {\{0,\ldots,k\}}^{d-2}$. Our method relies on generalizations of the "Gluing Lemma" of arXiv:1401.7130. A related result is that critical Bernoulli percolation on a slab satisfies the box-crossing property. Its proof is based on a new Russo-Seymour-Welsh type theorem for quasi-planar graphs. Thus, at criticality, the probability of an open path from $0$ of diameter $n$ decays polynomially in $n$. This strengthens the result of arXiv:1401.7130, where the absence of an infinite cluster at criticality was first established.

math.PR↗
arXiv2015

Monolayer Molybdenum Disulfide Nanoribbons with High Optical Anisotropy

Two-dimensional Molybdenum Disulfide (MoS2) has shown promising prospects for the next generation electronics and optoelectronics devices. The monolayer MoS2 can be patterned into quasi-one-dimensional anisotropic MoS2 nanoribbons (MNRs), in which theoretical calculations have predicted novel properties. However, little work has been carried out in the experimental exploration of MNRs with a width of less than 20 nm where the geometrical confinement can lead to interesting phenomenon. Here, we prepared MNRs with width between 5 nm to 15 nm by direct helium ion beam milling. High optical anisotropy of these MNRs is revealed by the systematic study of optical contrast and Raman spectroscopy. The Raman modes in MNRs show strong polarization dependence. Besides that the E' and A'1 peaks are broadened by the phonon-confinement effect, the modes corresponding to singularities of vibrational density of states are activated by edges. The peculiar polarization behavior of Raman modes can be explained by the anisotropy of light absorption in MNRs, which is evidenced by the polarized optical contrast. The study opens the possibility to explore quasione-dimensional materials with high optical anisotropy from isotropic 2D family of transition metal dichalcogenides.

cond-mat.mtrl-sci↗
arXiv2015

Large Scale Artificial Neural Network Training Using Multi-GPUs

This paper describes a method for accelerating large scale Artificial Neural Networks (ANN) training using multi-GPUs by reducing the forward and backward passes to matrix multiplication. We propose an out-of-core multi-GPU matrix multiplication and integrate the algorithm with the ANN training. The experiments demonstrate that our matrix multiplication algorithm achieves linear speedup on multiple inhomogeneous GPUs. The full paper of this project can be found at [1].

cs.DC↗
arXiv2015

A new framework for Euclidean summary statistics in the neural spike train space

Statistical analysis and inference on spike trains is one of the central topics in the neural coding. It is of great interest to understand the underlying structure of given neural data. Based on the metric distances between spike trains, recent investigations have introduced the notion of an average or prototype spike train to characterize the template pattern in the neural activity. However, as those metrics lack certain Euclidean properties, the defined averages are nonunique, and do not share the conventional properties of a mean. In this article, we propose a new framework to define the mean spike train where we adopt a Euclidean-like metric from an $L^p$ family. We demonstrate that this new mean spike train properly represents the average pattern in the conventional fashion, and can be effectively computed using a theoretically-proven convergent procedure. We compare this mean with other spike train averages and demonstrate its superiority. Furthermore, we apply the new framework in a recording from rodent geniculate ganglion, where background firing activity is a common issue for neural coding. We show that the proposed mean spike train can be utilized to remove the background noise and improve decoding performance.

q-bio.NC↗
arXiv2015

BLASX: A High Performance Level-3 BLAS Library for Heterogeneous Multi-GPU Computing

Basic Linear Algebra Subprograms (BLAS) are a set of low level linear algebra kernels widely adopted by applications involved with the deep learning and scientific computing. The massive and economic computing power brought forth by the emerging GPU architectures drives interest in implementation of compute-intensive level 3 BLAS on multi-GPU systems. In this paper, we investigate existing multi-GPU level 3 BLAS and present that 1) issues, such as the improper load balancing, inefficient communication, insufficient GPU stream level concurrency and data caching, impede current implementations from fully harnessing heterogeneous computing resources; 2) and the inter-GPU Peer-to-Peer(P2P) communication remains unexplored. We then present BLASX: a highly optimized multi-GPU level-3 BLAS. We adopt the concepts of algorithms-by-tiles treating a matrix tile as the basic data unit and operations on tiles as the basic task. Tasks are guided with a dynamic asynchronous runtime, which is cache and locality aware. The communication cost under BLASX becomes trivial as it perfectly overlaps communication and computation across multiple streams during asynchronous task progression. It also takes the current tile cache scheme one step further by proposing an innovative 2-level hierarchical tile cache, taking advantage of inter-GPU P2P communication. As a result, linear speedup is observable with BLASX under multi-GPU configurations; and the extensive benchmarks demonstrate that BLASX consistently outperforms the related leading industrial and academic projects such as cuBLAS-XT, SuperMatrix, MAGMA and PaRSEC.

cs.DC↗