检索结果-内蒙古大学图书馆

Lexicon Enhanced Chinese Sequence Labeling Using BERT Adapter

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Liu, Wei Fu, Xiyan Zhang, Yue Xiao, Wenming DAMO Academy Alibaba Group China College of Computer Science Nankai University China School of Engineering Westlake University China Institute of Advanced Technology Westlake Institute for Advanced Study China

Lexicon information and pre-trained models, such as BERT, have been combined to explore Chinese sequence labeling tasks due to their respective strengths. However, existing methods solely fuse lexicon features via a shallow and random initialized sequence layer and do not integrate them into the bottom layers of BERT. In this paper, we propose Lexicon Enhanced BERT (LEBERT) for Chinese sequence labeling, which integrates external lexicon knowledge into BERT layers directly by a Lexicon Adapter layer. Compared with existing methods, our model facilitates deep lexicon knowledge fusion at the lower layers of BERT. Experiments on ten Chinese datasets of three tasks including Named Entity Recognition, Word Segmentation, and Part-of-Speech Tagging, show that LEBERT achieves state-ofthe-art results. Copyright © 2021, The Authors. All rights reserved.

关键词： Computational linguistics

Magnetization and Spin Dynamics in Two-Dimensional Magnets

学校读者我要写书评

暂无评论

Magnetization and Spin Dynamics in Two-Dimensional Magnets

2023 IEEE International Magnetic Conference - Short Papers, INTERMAG Short Papers 2023

作者： Alahmed, Laith Nepal, Bhuwan Macy, Juan Casas, Brian Sapkota, Arjun Mazza, Alessandro R. Brahlek, Matthew Wen, Jiajia Jin, Wencan Zhang, Steven S.-L. Mewes, Claudia Zhang, Li-Chuan Mokrousov, Yuriy Zhang, Wei Lee, Young S. Balicas, Luis Mewes, Tim Zhang, Xiaoqian Li, Peng Department of Electrical and Computer Engineering Auburn University AuburnAL36849 United States Department of Physics University of Alabama TuscaloosaAL35487 United States National High Magnetic Field Laboratory Florida State University TallahasseeFL32310 United States Materials Science and Technology Division Oak Ridge National Laboratory Oak RidgeTN37831 United States Department of Physics Auburn University AuburnAL36849 United States Stanford Institute for Materials and Energy Sciences Slac National Accelerator Laboratory Menlo ParkCA94025 United States Department of Physics Case Western Reserve University ClevelandOH44106 United States Peter Grünberg Institut Institute for Advanced Simulation Forschungszentrum Jülich Jara Jülich52425 Germany Department of Physics and Astronomy University of North Carolina at Chapel Hill Chapel HillNC27599 United States Shenzhen Institute for Quantum Science and Engr. Southern University of Science and Technology Shenzhen518055 China School of Microelectronics University of Science and Technology of China Hefei230026 China Institute of Physics Johannes Gutenberg University Mainz Mainz55099 Germany

ISBN: (纸本)9798350338362

The discovery of magnetism in two-dimensional (2D) van der Waals (vdW) materials has flourished a new endeavor of fundamental problems in magnetism. In this talk, we will discuss the magnetization and spin dynamics in two types of 2D magnets: a room-temperature vdW magnetic crystal Fe5GeTe2 with a Curie temperature Tc = 332 K and Cu(1,3-bdc), a quasi-2D magnet with a Kagome lattice. Ferromagnetic resonance was performed with external magnetic fields applied along the c-axis (H||c) and the ab-plane (H||ab). We have found a sizable Landé g-factor difference between the H||c and H||ab cases for both Fe5GeTe2 and Cu(1,3-bdc). The FMR measurements have also revealed that Fe5GeTe2 has a damping constant comparable to Permalloy. Our experiments highlight key information regarding the magnetic state and spin scattering processes in 2D magnets, which can lead to implementations of 2D magnets based spintronic devices. © 2023 IEEE.

关键词： Nickel alloys

Dynamical causality under invisible confounders

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Yan, Jinling Zhang, Shao-Wu Zhang, Chihao Huang, Weitian Shi, Jifan Chen, Luonan MOE Key Laboratory of Information Fusion Technology School of Automation Northwestern Polytechnical University Xi’an710072 China Key Laboratory of Systems Biology Shanghai Institute of Biochemistry and Cell Biology Center for Excellence in Molecular Cell Science Chinese Academy of Sciences Shanghai200031 China NCMIS CEMS RCSDS Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing100190 China School of Mathematics Sciences University of Chinese Academy of Sciences Beijing100049 China School of Computer Science and Engineering South China University of Technology Guangdong Guangzhou510006 China Guangdong Institute of Intelligence Science and Technology Hengqin Guangdong Zhuhai519031 China Research Institute of Intelligent Complex Systems Fudan University Shanghai200433 China State Key Laboratory of Medical Neurobiology MOE Frontiers Center for Brain Science Institutes of Brain Science Fudan University Shanghai200032 China Shanghai Artificial Intelligence Laboratory Shanghai200232 China Key Laboratory of Systems Health Science of Zhejiang Province School of Life Science Hangzhou Institute for Advanced Study University of Chinese Academy of Sciences Chinese Academy of Sciences Hangzhou310024 China

Causality inference is prone to spurious causal interactions, due to the substantial confounders in a complex system. While many existing methods based on the statistical methods or dynamical methods attempt to address misidentification challenges, there remains a notable lack of effective methods to infer causality, in particular in the presence of invisible/unobservable confounders. As a result, accurately inferring causation with invisible confounders remains a largely unexplored and outstanding issue in data science and AI fields. In this work, we propose a method to overcome such challenges to infer dynamical causality under invisible confounders (CIC method) and further reconstruct the invisible confounders from time-series data by developing an orthogonal decomposition theorem in a delay embedding space. The core of our CIC method lies in its ability to decompose the observed variables not in their original space but in their delay embedding space into the common and private subspaces respectively, thereby quantifying causality between those variables both theoretically and computationally. This theoretical foundation ensures the causal detection for any high-dimensional system even with only two observed variables under many invisible confounders, which is actually a long-standing problem in the field. In addition to the invisible confounder problem, such a decomposition actually makes the intertwined variables separable in the embedding space, thus also solving the non-separability problem of causal inference. Extensive validation of the CIC method is carried out using various real datasets, and the experimental results demonstrates its effectiveness to reconstruct real biological networks even with unobserved confounders. © 2024, CC BY-NC-ND.

关键词： Time series

NaBiS2 as a Next-Generation Photovoltaic Absorber

学校读者我要写书评

暂无评论

NaBiS2 as a Next-Generation Photovoltaic Absorber

IEEE Conference Photovoltaic Specialists (PVSC)

作者： Yi-Teng Huang Seán R. Kavanagh Marcello Righetto Marin Rusu Igal Levine Thomas Unold Szymon J. Zelewski Alexander J. Sneyd Kaiwen Zhang Linjie Dai Andrew J. Britton Junzhi Ye Jaakko Julin Mari Napari Zhilong Zhang James Xiao Mikko Laitinen Laura Torrente-Murciano Samuel D. Stranks Akshay Rao Laura M. Herz David O. Scanlon Aron Walsh Robert L. Z. Hoye Cavendish Laboratory University of Cambridge Cambridge UK Department of Chemistry University College London London UK Department of Materials Imperial College London London UK Thomas Young Centre University College London London UK Department of Physics University of Oxford Clarendon Laboratory Oxford UK Helmholtz-Zentrum Berlin für Materialien und Energie GmbH Berlin Germany Department of Semiconductor Materials Engineering Faculty of Funda-mental Problems of Technology Wroclaw University of Science and Technology Wroclaw Poland Department of Chemical Engineering and Biotechnology University of Cambridge Cambridge UK School of Chemical and Process Engineering University of Leeds Leeds UK Department of Physics University of Jyvaskyla Jyvaskyla Finland School of Electronics and Computer Science University of Southampton Southampton UK Institute for Advanced Study Technical University of Munich Garching Germany

ISBN: (数字)9781665464260

ISBN: (纸本)9781665475822

$\mathbf{I}-\mathbf{V}-\mathbf{VI}_{2}$ ternary chalcogenides were gaining more attention recently owing to their high air stability along with earth-abundant and nontoxic compounds. More interestingly, some of them have been predicted to be potential “perovskite-inspired materials” that can exhibit defect tolerance. In this work, we investigated the optoelectronic properties of cation-disordered $\mathbf{NaBiS}_{2}$ nanocrystals, which have a steep absorption onset, with absorption coefficients $> \mathbf{10}^{5} \ \mathbf{cm}^{-1}$ just above its pseudo-direct bandgap of 1.4 eVe Surprisingly, an ultrafast photoconductivity decay and long-lived charge-carrier population almost independent of defect densities were simultaneously observed in N aBiS2 nanocrystals. Through the density functional theory calculations, these unusual features are found to be highly associated with the non-bonding $\mathbf{S} \ {p}$ orbitals above the upper valence band, which can be further enhanced by cation inhomogeneity. Our work hence reveals the critical role of cation disorder on both absorption characteristics and charge-carrier kinetics of $\mathbf{I}-\mathbf{V}-\mathbf{VI}_{2}$ ternary chalcogenides.

关键词： Photovoltaic systems Couplings Chalcogenides Absorption Nanocrystals Photonic band gap Orbits Optical materials Ultrafast optics Photoconductivity

Multi-modal isolated sign language recognition based on self-paced learning

学校读者我要写书评

暂无评论

Expert Systems with Applications 2025年 291卷

作者： Yazhou Ren Hongkai Li Yuhao Li Jingyu Pu Xiaorong Pu Siyuan Jing Peng Jin Lifang He Shenzhen Institute for Advanced Study University of Electronic Science and Technology of China Shenzhen Guangdong China School of Computer Science and Engineering University of Electronic Science and Technology of China Chengdu Sichuan China Sichuan Provincial Key Lab of Philosophy and Social Science for Language Intelligence in Special Education Leshan Normal University Leshan Sichuan China Department of Computer Science and Engineering Lehigh University Bethlehem PA USA

Sign language is an essential communication tool for the deaf and hard-of-hearing community, yet its complexity poses a significant barrier to widespread adoption. Effective Sign Language Recognition (SLR) systems are crucial for enhancing communication between deaf and hearing individuals. This paper introduces a multi-modal SLR framework that integrates self-paced learning into SLR for the first time. By leveraging transformer and graph convolutional network technologies in an ensemble fusion, it significantly enhances accuracy without increasing computational complexity. Another major challenge in SLR system development is the regional variation and the lack of comprehensive, diverse datasets. To address this, we introduce the ISW-1000 dataset, featuring 10,000 video samples of 1000 isolated Chinese sign words, performed by 10 deaf and hard-of-hearing signers in a uniform environment. This dataset provides a valuable resource for SLR research, encompassing multiple data modalities such as RGB frames, optical flow, and skeleton. We have established benchmark training and testing sets to support both signer-independent and general model evaluations. Our extensive experiments on the ISW-1000 dataset show that our multi-modal self-paced approach significantly surpasses existing single-modal and multi-modal methods in recognition accuracy. Both the code and the proposed ISW-1000 dataset are publicly available at https://***/leehkk/Chinese-SLR .

关键词：

Exploring the Efficacy of Automatically Generated Counterfactuals for Sentiment Analysis

学校读者我要写书评

暂无评论

arXiv 2021年

作者： Yang, Linyi Li, Jiazheng Cunningham, Pádraig Zhang, Yue Smyth, Barry Dong, Ruihai The Insight Centre for Data Analytics University College Dublin Ireland School of Computer Science University College Dublin Ireland School of Engineering Westlake University China Institute of Advanced Technology Westlake Institute for Advanced Study China

While state-of-the-art NLP models have been achieving the excellent performance of a wide range of tasks in recent years, important questions are being raised about their robustness and their underlying sensitivity to systematic biases that may exist in their training and test data. Such issues come to be manifest in performance problems when faced with out-of-distribution data in the field. One recent solution has been to use counterfactually augmented datasets in order to reduce any reliance on spurious patterns that may exist in the original data. Producing high-quality augmented data can be costly and time-consuming as it usually needs to involve human feedback and crowdsourcing efforts. In this work, we propose an alternative by describing and evaluating an approach to automatically generating counterfactual data for the purpose of data augmentation and explanation. A comprehensive evaluation on several different datasets and using a variety of state-of-the-art benchmarks demonstrate how our approach can achieve significant improvements in model performance when compared to models training on the original data and even when compared to models trained with the benefit of human-generated augmented data. © 2021, CC BY-NC-SA.

关键词： Sentiment analysis

Ferroelectrets:Heterogenous polymer electrets with high piezoelectric sensitivity for transducers

学校读者我要写书评

暂无评论

Journal of Advanced Dielectrics 2023年第4期13卷 3-26页

作者： Xunlin Qiu Peng Fang Axel Mellinger Ruy Alberto Pisani Altafim Werner Wirges Gunnar Gidion Dmitry Rychkov Shanghai Key Laboratory of Intelligent Sensing and Detection Technology School of Mechanical and Power Engineering East China University of Science and Technology Shanghai 200237P.R.China CAS Key Laboratory of Human-Machine Intelligence-Synergy Systems Shenzhen Institutes of Advanced Technology Shenzhen 518055P.R.China Department of Physics Central Michigan University 223 Dow Science ComplexMount PleasantMI 48859USA Computer Systems Department Informatics Center Federal University of ParáıbaJoão Pessoa-PBBrazil Department of Physics and Astronomy University of Potsdam PotsdamBrandenburgGermany Laboratory for Electrical Instrumentation Department of Microsystems Engineering-IMTEK University of FreiburgGeorges-Köhler-Allee 106Freiburg 79110Germany Technology and Study Center Weissenburg Deggendorf Institute of Technology WeissenburgBavariaGermany

Nowadays,the demand for advanced functional materials in transducer technology is growing *** materials transform mechanical variables(displacement or force)into electrical signals(charge or voltage)and vice *** are interesting from both fundamental and application points of ***(also called piezoelectrets)are a relatively young group of piezo-,pyro-and ferroelectric *** exhibit ferroic behavior phenomenologically undistinguishable from that of traditional ferroelectrics,although the materials per se are essentially non-polar space-charge electrets with artificial macroscopic dipoles(i.e.,internally charged cavities).A lot of work has been done on ferroelectrets and their applications up to *** this paper,we review and discuss mostly the work done at University of Potsdam on the research and development of *** will,however,also mention important results from other teams,and prospect the challenges and future progress trend of the field of ferroelectret research.

关键词： Ferroelectret piezoelectret space-charge electret transducer advanced functional materials

Magnon corner states in twisted bilayer honeycomb magnets

学校读者我要写书评

暂无评论

arXiv 2022年

作者： Hua, Chun-Bo Xiao, Feiping Liu, Zheng-Rong Sun, Jin-Hua Gao, Jin-Hua Chen, Chui-Zhen Tong, Qingjun Zhou, Bin Xu, Dong-Hui School of Electronic and Information Engineering Hubei University of Science and Technology Xianning437100 China Laboratory of Optoelectronic Information and Intelligent Control Hubei University of Science and Technology Xianning437100 China School of Physics and Electronics Hunan University Changsha410082 China Department of Physics Hubei University Wuhan430062 China Department of Physics Ningbo University Ningbo315211 China School of Physics Wuhan National High Magnetic Field Center Huazhong University of Science and Technology Wuhan430074 China Institute for Advanced Study School of Physical Science and Technology Soochow University Suzhou215006 China Department of Physics Chongqing Key Laboratory for Strongly Coupled Physics Chongqing University Chongqing400044 China Center of Quantum Materials and Devices Chongqing University Chongqing400044 China

The study of symmetry-protected topological phases of matter has been extended from fermionic electron systems to various bosonic systems. Bosonic topological magnon phases in magnetic materials have received much attention because of their exotic uncharged topologically protected boundary modes and the potential for dissipationless magnonics and spintronic applications. Here, we establish twisted bilayer honeycomb magnets as a platform for hosting second-order topological magnon insulators (SOTMIs) without fine-tuning. We employ a simple, minimal Heisenberg spin model to describe misaligned bilayer sheets of honeycomb ferromagnetic magnets with a large commensurate twist angle. We found that the higher-order topology in this bilayer system shows a significant dependence on the interlayer exchange coupling. The SOTMI, featuring topologically protected magnon corner states that go beyond the conventional bulk-boundary correspondence, appears for ferromagnetic interlayer couplings, while the twisted bilayer exhibits a nodal phase in the case of antiferromagnetic interlayer coupling. At last, relevance to twisted bilayer CrI3 is also discussed. © 2022, CC BY.

关键词： Chromium compounds

Beyond the band edge: Unveiling high-mobility hot carriers in a two-dimensional conjugated coordination polymer

学校读者我要写书评

暂无评论

arXiv 2025年

作者： Fu, Shuai Huang, Xing Gao, Guoquan Petkov, Petko St. Gao, Wenpei Zhang, Jianjun Gao, Lei Zhang, Heng Liu, Min Hambsch, Mike Zhang, Wenjie Zhang, Jiaxu Li, Keming Kaiser, Ute Parkin, Stuart S.P. Mannsfeld, Stefan C.B. Zhu, Tong Wang, Hai I. Wang, Zhiyong Dong, Renhao Feng, Xinliang Bonn, Mischa Center for Advancing Electronics Dresden Faculty of Chemistry and Food Chemistry Technische Universität Dresden Dresden Germany Max Planck Institute for Polymer Research Mainz Germany Laser Micro/Nano Fabrication Laboratory School of Mechanical Engineering Beijing Institute of Technology Beijing China Faculty of Chemistry and Pharmacy University of Sofia Sofia Bulgaria State Key Laboratory of Metal Matrix Composites School of Materials Science and Engineering Future Material Innovation Center Zhangjiang Institute for Advanced Study Shanghai Jiao Tong University Shanghai China Center for Advancing Electronics Dresden Faculty of Electrical and Computer Engineering TUD Dresden University of Technology Germany Germany Central Facility for Materials Science Electron Microscopy Universität Ulm Ulm Germany Nanophotonics Debye Institute for Nanomaterials Science Utrecht University Utrecht Netherlands Department of Chemistry The University of Hong Kong 999077 Hong Kong HKU-SIRI Shenzhen518048 China

Hot carriers, inheriting excess kinetic energy from high-energy photons, underpin numerous optoelectronic applications involving non-equilibrium transport processes. Current research on hot carriers has predominantly focused on inorganic materials, with little attention paid to organic-based systems due to their ultrafast energy relaxation and inefficient charge transport. Here, we overturn this paradigm by demonstrating highly mobile hot carriers in solution-processable, highly crystalline two-dimensional conjugated coordination polymer (2D c-CP) Cu3BHT (BHT = benzenehexathiol) films. Leveraging a suite of ultrafast spectroscopic and imaging techniques, we unravel the microscopic charge transport landscape in Cu3BHT films following non-equilibrium photoexcitation across temporal, spatial, and frequency domains, revealing two distinct high-mobility transport regimes. In the non-equilibrium transport regime, hot carriers achieve ultrahigh mobility of ~2,000 cm2 V–1 s–1, traversing grain boundaries up to ~300 nm within a picosecond. In the quasi-equilibrium transport regime, free carriers exhibit Drude-type band-like transport with a remarkable mobility of ~400 cm2 V–1 s–1 and an intrinsic diffusion length exceeding 1 µm. These findings establish 2D c-CPs as versatile platforms for exploring high-mobility non-equilibrium transport, unlocking new opportunities for organic-based hot carrier applications. Copyright © 2025, The Authors. All rights reserved.

关键词： Kinetic energy