检索结果-内蒙古大学图书馆

IMPACT: Importance-Informed Prefetching and Caching for I/O-Bound DNN Training

IEEE Transactions on Computers 2025年

作者： Chen, Weijian He, Shuibing Zhang, Ruidong Zhang, Xuechen Chen, Ping Yang, Siling Qu, Haoyang Zhan, Xuan Zhejiang University State Key Laboratory of Blockchain and Data Security China Institute of Blockchain and Data Security Hangzhou China Zhejiang Key Laboratory of Big Data Intelligent Computing China Washington State University Vancouver School of Engineering and Computer Science VancouverWA98686 United States

Fetching large amounts of DNN training data from storage systems causes high I/O latency and GPU stalls. Importance sampling can reduce data processing on GPUs while maintaining model accuracy, but current frameworks lack a prefetching and caching layer to optimize data fetches and cache management based on sample importance. This leads to unnecessary fetches, poor cache hit ratios, and random I/Os. We present ImPACT, an importance-informed prefetching and caching system, to accelerate I/O-bound DNN training. First, we propose an importance-informed prefetching technique to reduce the prefetching of unimportant data. Then, we introduce an importance-aware caching layer, partitioned into two regions: H-cache and L-cache, which store samples of high importance and low importance respectively. Rather than using recency or frequency, we manage data items in H-cache according to their corresponding sample importance. When there is a cache miss in L-cache, we use sample substitutability and dynamic packaging to improve the cache hit ratio and reduce the number of random I/Os. Our experimental results show that ImPACT has a negligible impact on training accuracy while speeding up DNN training by up to 3.5× compared to state-of-the-art prefetching and caching systems. © 1968-2012 IEEE.

关键词： Cache memory

来源：评论

学校读者我要写书评

暂无评论

Microstructure evolution and ablation behavior of C/C-HfC-ZrC-SiC composites in extreme laser ablation environment

引用

Journal of Alloys and Compounds 2025年 1022卷

作者： Zhang, Menglin Li, Xiaoxuan Yao, Xiyuan Hu, Dou Fu, Qiangang State Key Laboratory of Solidification Processing Shaanxi Key Laboratory of Fiber Reinforced Light-Weight Composites Northwestern Polytechnical University Shaanxi Xi'an710072 China Department of Data and Systems Engineering University of Hong Kong 999077 Hong Kong

To balance the excellent protection performance, lightweight design and low preparation cost of carbon-based composites for extreme high-temperature environment, the C/C-HfC-ZrC-SiC composites with varying Hf/Zr molar ratios via reactive melt infiltration were optimized under high-energy laser ablation condition (39.8 MW/m2). Combined with ablation recession simulations, the increase in Hf/Zr ratio facilitates the improvement of the laser ablation resistance, while the performance enhancement effect at 40 s ablation is not apparent. The higher structural stability of (Hf,Zr)O2 solid solution enhances the ablation resistance of the equimolar Hf/Zr ratio composites, with the linear variation rate of only 3.83 μm/s. The thermal stress mismatch between the carbon fiber bundles and the ceramic-rich layer is susceptible to crack initiation and extension, resulting in destructive sublimation and oxidation. In consideration of the overall density, ablation performance and cost, the equimolar Hf/Zr ratio composites exhibit the best performance during laser ablation. © 2025 Elsevier B.V.

关键词： Laser ablation

来源：评论

学校读者我要写书评

暂无评论

CroPrompt: Cross-task Interactive Prompting for Zero-shot Spoken Language Understanding

CroPrompt: Cross-task Interactive Prompting for Zero-shot Sp...

引用

2025 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2025

作者： Qin, Libo Wei, Fuxuan Chen, Qiguang Zhou, Jingxuan Huang, Shijue Si, Jiasheng Lu, Wenpeng Che, Wanxiang School of Computer Science and Engineering Central South University China Key Laboratory of Data Intelligence and Advanced Computing in Provincial Universities Soochow University China Research Center for Social Computing and Information Retrieval Harbin Institute of Technology China Harbin Institute of Technology Shenzhen China Key Laboratory of Computing Power Network and Information Security Ministry of Education Qilu University of Technology Shandong Academy of Sciences China

ISBN: (纸本)9798350368741

Slot filling and intent detection are two highly correlated tasks in spoken language understanding (SLU). Recent SLU research attempts to explore zero-shot prompting techniques in large language models to alleviate the data scarcity problem. Nevertheless, the existing prompting work ignores the cross-task interaction information for SLU, which leads to sub-optimal performance. To solve this problem, we present the pioneering work of Cross-task Interactive Prompting (CroPrompt) for SLU, which enables the model to interactively leverage the information exchange across the correlated tasks in SLU. Additionally, we further introduce a multi-task self-consistency mechanism to mitigate the error propagation caused by the intent information injection. We conduct extensive experiments on the standard SLU benchmark and the results reveal that CroPrompt consistently outperforms the existing prompting approaches. In addition, the multi-task self-consistency mechanism can effectively ease the error propagation issue, thereby enhancing the performance. We hope this work can inspire more research on cross-task prompting for SLU. © 2025 IEEE.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Distinguishing LLM-Generated from Human-Written Code by Contrastive Learning

引用

ACM Transactions on Software engineering and Methodology 2025年第4期34卷

作者： Xu, Xiaodan Ni, Chao Guo, Xinrong Liu, Shaoxuan Wang, Xiaoya Liu, Kui Yang, Xiaohu State Key Laboratory of Blockchain and Data Security Zhejiang University Hangzhou China Software Engineering Application Technology Lab Hangzhou Huawei China Institute of Blockchain and Data Security China

Large language models (LLMs), such as ChatGPT released by OpenAI, have attracted significant attention from both industry and academia due to their demonstrated ability to generate high-quality content for various tasks. Despite the impressive capabilities of LLMs, there are growing concerns regarding their potential risks in various fields, such as news, education, and software engineering. Recently, several commercial and open source LLM-generated content detectors have been proposed, which, however, are primarily designed for detecting natural language content without considering the specific characteristics of program code. This article aims to fill this gap by proposing a novel ChatGPT-generated code detector, CodeGPTSensor, based on a contrastive learning framework and a semantic encoder built with UniXcoder. To assess the effectiveness of CodeGPTSensor on differentiating ChatGPT-generated code from human-written code, we first curate a large-scale Human and Machine comparison Corpus (HMCorp), which includes 550k pairs of human-written and ChatGPT-generated code (i.e., 288k Python code pairs and 222k Java code pairs). Based on the HMCorp dataset, our qualitative and quantitative analysis of the characteristics of ChatGPT-generated code reveals the challenge and opportunity of distinguishing ChatGPT-generated code from human-written code with their representative features. Our experimental results indicate that CodeGPTSensor can effectively identify ChatGPT-generated code, outperforming all selected baselines. © 2025 Copyright held by the owner/author(s). Publication rights licensed to ACM.

关键词： Open source software

来源：评论

学校读者我要写书评

暂无评论

NSPG-Miner: Mining Repetitive Negative Sequential Patterns

arXiv

引用

arXiv 2025年

作者： Li, Yan Wang, Zhulin Liu, Jing Guo, Lei Fournier-Viger, Philippe Wu, Youxi Wu, Xindong School of Economics and Management Hebei University of Technology Tianjin300401 China School of Artificial Intelligence Hebei University of Technology Tianjin300401 China State Key Laboratory of Reliability and Intelligence of Electrical Equipment Hebei University of Technology Tianjin300401 China College of Computer Science and Software Engineering Shenzhen University Shenzhen518061 China Hebei Key Laboratory of Big Data Computing 300401 China Key Laboratory of Knowledge Engineering with Big Data The Ministry of Education of China Hefei University of Technology Hefei230009 China

Sequential pattern mining (SPM) with gap constraints (or repetitive SPM or tandem repeat discovery in bioinformatics) can find frequent repetitive subsequences satisfying gap constraints, which are called positive sequential patterns with gap constraints (PSPGs). However, classical SPM with gap constraints cannot find the frequent missing items in the PSPGs. To tackle this issue, this paper explores negative sequential patterns with gap constraints (NSPGs). We propose an efficient NSPG-Miner algorithm that can mine both frequent PSPGs and NSPGs simultaneously. To effectively reduce candidate patterns, we propose a pattern join strategy with negative patterns which can generate both positive and negative candidate patterns at the same time. To calculate the support (frequency of occurrence) of a pattern in each sequence, we explore a NegPair algorithm that employs a key-value pair array structure to deal with the gap constraints and the negative items simultaneously and can avoid redundant rescanning of the original sequence, thus improving the efficiency of the algorithm. To report the performance of NSPG-Miner, 11 competitive algorithms and 11 datasets are employed. The experimental results not only validate the effectiveness of the strategies adopted by NSPG-Miner, but also verify that NSPG-Miner can discover more valuable information than the state-ofthe-art algorithms. Algorithms and datasets can be downloaded from https://***/wuc567/PatternMining/tree/master/NSPG-Miner. Copyright © 2025, The Authors. All rights reserved.

关键词： Miners

来源：评论

学校读者我要写书评

暂无评论

NUMERICAL ENERGY DISSIPATION FOR TIME-FRACTIONAL PHASE-FIELD EQUATIONS

引用

Journal of Computational Mathematics 2025年第3期43卷 515-539页

作者： Chaoyu Quan Tao Tang Jiang Yang School of Science and Engineering The Chinese University of Hong KongShenzhen 518172China Division of Science and Technology BNU-HKBU United International CollegeZhuhai 519087China Guangdong Provincial Key Laboratory of Interdisciplinary Research and Application for Data Science BNU-HKBU United International CollegeZhuhai 519087China Department of Mathematics SUSTech International Center for Mathematics&National Center for Applied Mathematics Shenzhen(NCAMS)Southern University of Science and TechnologyShenzhen 518055China

The numerical integration of phase-field equations is a delicate task which needs to recover at the discrete level intrinsic properties of the solution such as energy dissipation and maximum *** the theory of energy dissipation for classical phase field models is well established,the corresponding theory for time-fractional phase-field models is still *** this article,we study certain nonlocal-in-time energies using the first-order stabilized semi-implicit L1 *** particular,we will establish a discrete fractional energy law and a discrete weighted energy *** extension for a(2-α)-order L1 scalar auxiliary variable scheme will be ***,we demonstrate that the energy bound is preserved for the L1 schemes with nonuniform time *** numerical experiments are carried to verify our theoretical analysis.

关键词： Time-fractional phased-field equation Allen-Cahn equations Cahn-Hilliard equations Caputo fractional derivative Energy dissipation

来源：评论

学校读者我要写书评

暂无评论

Extracting High-order Connectivity of EEG-based Dynamic Functional Connectivity Networks for Diagnosis of Major Depressive Disorder 24

Extracting High-order Connectivity of EEG-based Dynamic Func...

引用

4th Asia-Pacific Artificial Intelligence and Big data Forum, AIBDF 2024

作者： Zhao, Feng Fan, Wenxuan Han, Zhongwei Chen, Hongyu School of Computer Science and Technology Shandong Technology and Business University Shandong Yantai China Yantai Key Laboratory of Big Data Modeling and Intelligent Computing Shandong Yantai China Immersion Technology and Evaluation Shandong Engineering Research Center Shandong Yantai China Information Engineering College Yantai Institute of Technology Shandong Yantai China Shandong Technology and Business University Shandong Yantai China

ISBN: (纸本)9798400710865

Functional connectivity networks (FCNs), particularly those utilizing the phase lag index (PLI) method, have been instrumental in elucidating the pathological features of Major Depressive Disorder (MDD) by assessing the synchrony and lag of electroencephalogram (EEG) signal waveforms. Currently, the more prevalent methods are traditional FCNs that quantify phase synchronization as scalar values and dynamic FCNs (D-FCNs) that employ sliding window approaches to capture the instantaneous variations in functional connectivity. However, these two methods merely capture the low-level functional connections between paired regions of interest (ROIs) from both the static and dynamic perspectives, and fail to reflect the high-level functional connections involving multiple ROIs. Moreover, the feature dimension of the dynamic FCN is excessively high, and there exists a phase mismatch issue between different samples, which renders it impossible to be directly applied to the classification task, and no better solution is available at present. To mitigate these issues, we proposed a comprehensive framework that includes segmenting EEG sequences based on PLI to construct dynamic FCNs;employing seventh-order central moments for feature extraction to form low-order FCNs (Lo-FCNs);and establishing high-order FCNs (Ho-FCNs) based on Pearson correlation theory derived from Lo-FCNs. We conducted classification experiments across four distinct EEG frequency bands using our proposed framework while analyzing relevant features through graph theory and brain topological mapping techniques. The experiments showed that the proposed framework has a high classification accuracy rate in the classification experiments of MDD and NC, and proved the importance of high-order features in classification experiments, indicating that they may serve as physiological markers for MDD diagnosis. © 2024 Copyright held by the owner/author(s).

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

SwimVG: Step-wise Multimodal Fusion and Adaption for Visual Grounding

arXiv

引用

arXiv 2025年

作者： Shi, Liangtao Liu, Ting Hu, Xiantao Hu, Yue Yin, Quanjun Hong, Richang Key Laboratory of Knowledge Engineering with Big Data Hefei University of Technology Hefei230009 China Ministry of Education and School of Computer Science and Information Engineering Hefei University of Technology Hefei230009 China School of systems engineering National University of Defense Technology Hunan Province Changsha410073 China Department of Computer Science and Engineering Nanjing University of Science and Technology Nanjing210014 China

Visual grounding aims to ground an image region through natural language, which heavily relies on cross-modal alignment. Most existing methods transfer visual/linguistic knowledge separately by fully fine-tuning uni-modal pre-trained models, followed by a simple stack of visual-language transformers for multimodal fusion. However, these approaches not only limit adequate interaction between visual and linguistic contexts, but also incur significant computational costs. Therefore, to address these issues, we explore a step-wise multimodal fusion and adaption framework, namely SwimVG. Specifically, SwimVG proposes step-wise multimodal prompts (Swip) and cross-modal interactive adapters (CIA) for visual grounding, replacing the cumbersome transformer stacks for multimodal fusion. Swip can improve the alignment between the vision and language representations step by step, in a token-level fusion manner. In addition, weight-level CIA further promotes multimodal fusion by cross-modal interaction. Swip and CIA are both parameter-efficient paradigms, and they fuse the cross-modal features from shallow to deep layers gradually. Experimental results on four widely-used benchmarks demonstrate that SwimVG achieves remarkable abilities and considerable benefits in terms of efficiency. Our code is available at https://***/liuting20/SwimVG. Copyright © 2025, The Authors. All rights reserved.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

MCITEBENCH: A Benchmark for Multimodal Citation Text Generation in MLLMs

arXiv

引用

arXiv 2025年

作者： Hu, Caiyu Zhang, Yikai Zhu, Tinghui Ye, Yiwei Xiao, Yanghua Shanghai Key Laboratory of Data Science School of Computer Science Fudan University China School of Computer Engineering and Science Shanghai University China

Multimodal Large Language Models (MLLMs) have advanced in integrating diverse modalities but frequently suffer from hallucination. A promising solution to mitigate this issue is to generate text with citations, providing a transparent chain for verification. However, existing work primarily focuses on generating citations for text-only content, overlooking the challenges and opportunities of multimodal contexts. To address this gap, we introduce MCITEBENCH, the first benchmark designed to evaluate and analyze the multimodal citation text generation ability of MLLMs. Our benchmark comprises data derived from academic papers and review-rebuttal interactions, featuring diverse information sources and multimodal content. We comprehensively evaluate models from multiple dimensions, including citation quality, source reliability, and answer accuracy. Through extensive experiments, we observe that MLLMs struggle with multimodal citation text generation. We also conduct deep analyses of models’ performance, revealing that the bottleneck lies in attributing the correct sources rather than understanding the multimodal content. Copyright © 2025, The Authors. All rights reserved.

关键词： Benchmarking

来源：评论

学校读者我要写书评

暂无评论

RobNODDI: Robust NODDI Parameter Estimation with Adaptive Sampling Under Continuous Representation 15th

RobNODDI: Robust NODDI Parameter Estimation with Adaptive S...

引用

15th International Workshop on Computational Diffusion MRI, CDMRI 2024, held in conjunction with 27th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2024

作者： Xiao, Taohui Cheng, Jian Fan, Wenxin Yang, Jing Li, Cheng Dong, Enqing Wang, Shanshan School of Mechanical Electrical and Information Engineering Shandong University Weihai264209 China State Key Laboratory of Software Development Environment Beihang University Beijing China Key Laboratory of Data Science and Intelligent Computing Institute of International Innovation Beihang University Zhejiang Hangzhou China Paul C. Lauterbur Research Center for Biomedical Imaging Shenzhen Institute of Advanced Technology Chinese Academy of Sciences Guangdong Shenzhen China Peng Cheng Laboratory Guangdong Shenzhen China

ISBN: (纸本)9783031869198

Neurite Orientation Dispersion and Density Imaging (NODDI) is an important imaging technology used to evaluate the microstructure of brain tissue, which is of great significance for the discovery and treatment of various neurological diseases. Current deep learning-based methods perform parameter estimation through diffusion magnetic resonance imaging (dMRI) with a small number of diffusion gradients. These methods speed up parameter estimation and improve accuracy. However, the diffusion directions used by most existing deep learning models during testing needs to be strictly consistent with the diffusion directions during training. This results in poor generalization and robustness of deep learning models in dMRI parameter estimation. In this work, we first verify that the parameter estimation performance of current mainstream methods will significantly decrease when the testing diffusion directions and the training diffusion directions are inconsistent. A robust NODDI parameter estimation method with adaptive sampling under continuous representation (RobNODDI) is proposed. Furthermore, long short-term memory (LSTM) units and fully connected layers are selected to learn continuous representation signals. To this end, we use a total of 100 subjects to conduct experiments based on the Human Connectome Project (HCP) dataset, of which 60 are used for training, 20 are used for validation, and 20 are used for testing. The test results indicate that RobNODDI improves the generalization performance and robustness of the deep learning model, enhancing the stability and flexibility of deep learning NODDI parameter estimatimation applications. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：