检索结果-内蒙古大学图书馆

A prompt-based approach to adversarial example generation and robustness enhancement

Frontiers of computer Science 2024年第4期18卷 85-96页

作者： Yuting YANG Pei HUANG Juan CAO Jintao LI Yun LIN Feifei MA Key Lab of Intelligent Information Processing of Chinese Academy of Sciences(CAS) Institute of Computing TechnologyCASBeijing 100190China School of Computer Science and Technology University of Chinese Academy of SciencesBeijing 100049China Department of Computer Science Stanford UniversityCA 94305USA School of Computing National University of SingaporeSingapore 119077Singapore Laboratory of Parallel Software and Computational Science Institute of SoftwareChinese Academy of SciencesBeijing 100190China

Recent years have seen the wide application of natural language processing(NLP)models in crucial areas such as finance,medical treatment,and news media,raising concerns about the model robustness and *** find that prompt paradigm can probe special robust defects of pre-trained language *** prompt texts are first constructed for inputs and a pre-trained language model can generate adversarial examples for victim models via *** results show that prompt paradigm can efficiently generate more diverse adversarial examples besides synonym ***,we propose a novel robust training approach based on prompt paradigm which incorporates prompt texts as the alternatives to adversarial examples and enhances robustness under a lightweight minimax-style optimization *** on three real-world tasks and two deep neural models show that our approach can significantly improve the robustness of models to resist adversarial attacks.

关键词： robustness adversarial example prompt learning pre-trained language model

来源：评论

学校读者我要写书评

暂无评论

The Design of a computing Platform architecture for Collaborative Task Processing in Avionics System 7th

The Design of a Computing Platform Architecture for Collabor...

引用

7th China Aeronautical Science and technology Conference, CASTC 2024

作者： Lu, Chenxin Xie, Jianchun Liu, Zuolong Xi’an Aeronautics Computing Technique Research Institute AVIC Xi’an China Key Laboratory of Science and Technology on Airborne and Missileborne Computer Xi’an China

ISBN: (纸本)9789819797738

With the rapid development and application of the concept of multi-aircraft coordinated operations, there are higher requirements for the performance of airborne computing platforms in processing multi-aircraft coordinated tasks. In this study, we have examined the architecture and computing mode of airborne computing platforms in distributed coordinated application scenarios. We have optimized traditional airborne embedded computer architecture, integrated cloud-edge computing technology and heterogeneous parallel computing technology to enhance the performance of airborne computing platforms. We proposes a multi-level heterogeneous distributed collaborative computing platform architecture. The hardware provides modular computing units, while the software implements distributed collaborative resource management and task management mechanisms, along with a unified high-speed network. Additionally, to meet real-time processing requirements for massive data in coordinated task scenarios, we propose a high-throughput computing mode based on the defined platform architecture to ensure task execution throughput through massively parallel acceleration arrays. By integrating the proposed distributed computing architecture with the high throughput computing model, we have developed an integrated airborne distributed collaborative computing platform. In conclusion, we have conducted an analysis of the application of this platform in typical task scenarios of formation coordinated combat. © Chinese Society of Aeronautics and Astronautics 2025.

关键词： HIgh speed networks

来源：评论

学校读者我要写书评

暂无评论

Passive on-chip isolators based on the thin-film lithium niobate platform

引用

Chinese Physics B 2025年第3期34卷 408-413页

作者： Jiacheng Liu Gongyu Xia Qilin Hong Pingyu Zhu Kai-Kai Zhang keyu Xia Ping Xu Shiqiao Qin Zhihong Zhu College of Advanced Interdisciplinary Studies&Hunan Provincial Key Laboratory of Novel Nano Optoelectronic Information Materials and Devices National University of Defense TechnologyChangsha 410073China Institute for Quantum Information and State Key Laboratory of High-Performance Computing College of Computer Science and TechnologyNational University of Defense TechnologyChangsha 410073China College of Engineering and Applied Sciences National Laboratory of Solid State Microstructuresand Collaborative Innovation Center of Advanced MicrostructuresNanjing UniversityNanjing 210023China

Optical isolators,the photonic analogs of electronic diodes,are essential for ensuring the unidirectional flow of light in optical systems,thereby mitigating the destabilizing effects of back ***-film lithium niobate(TFLN),hailed as“the silicon of photonics,”has emerged as a pivotal material in the realm of chip-scale nonlinear optics,propelling the demand for compact optical *** report a breakthrough in the development of a fully passive,integrated optical isolator on the TFLN platform,leveraging the Kerr effect to achieve an impressive 10.3 dB of isolation with a minimal insertion loss of 1.87 *** theoretical simulations have demonstrated that our design,when applied to a microring resonator with a Q factor of 5×10^(6),can achieve 20 dB of isolation with an input power of merely 8 *** advancement underscores the immense potential of lithium niobate-based Kerr-effect isolators in propelling the integration and application of high-performance on-chip lasers,heralding a new era in integrated photonics.

关键词： thin-film lithium niobate Kerr effect optical isolator

来源：评论

学校读者我要写书评

暂无评论

Heralded path-entangled NOON states generation from a reconfigurable photonic chip

引用

Chinese Physics B 2022年第6期31卷 358-362页

作者： Xinyao Yu Pingyu Zhu Yang Wang Miaomiao Yu Chao Wu Shichuan Xue Qilin Zheng Yingwen Liu Junjie Wu Ping Xu Institute for Quantum Information&State Key Laboratory of High Performance Computing College of Computer Science and TechnologyNational University of Defense TechnologyChangsha 410073China

Maximal multi-photon entangled states,known as NOON states,play an essential role in quantum *** the number of photons growing,NOON states are becoming increasingly powerful and advantageous for obtaining supersensitive and super-resolved *** this paper,we propose a universal scheme for generating three-and four-photon path-entangled NOON states on a reconfigurable photonic chip via photons subtracted from pairs and detected by heralding *** method is postselection free,enabling phase supersensitive measurements and sensing at the Heisenberg *** NOON-state generator allows for integration of quantum light sources as well as practical and portable precision phase-related measurements.

关键词： path-entangled NOON state phase measurement

来源：评论

学校读者我要写书评

暂无评论

Stochastic Task Offloading Problems for Edge computing 17th

Stochastic Task Offloading Problems for Edge Computing

引用

17th CCF Conference on computer Supported Cooperative Work and Social computing, ChineseCSCW 2022

作者： Ding, Kexin Zhong, Zhi Zhu, Jie Nanjing University of Posts and Telecommunications Nanjing210003 China Institute of Computing Technology Academy of Sciences State Key Laboratory Chinese of Computer Architecture Beijing100864 China

ISBN: (纸本)9789819923557

The edge-cloud computing is extensively deployed to provide convenient computing. A stochastic task offloading problem in edge-cloud environment is considered. For the problem under study, a greedy simulated annealing heuristic algorithm based online offloading framework (SAOF) is proposed. In consideration of the stochastic arrival of independent tasks and the heterogeneity of resources, SAOF offloads tasks to cloud servers or edge servers dynamically. During the offloading process, each server preemptively schedules the uncompleted tasks according to the latency-sensitivity of the task. The optimization objective is to minimize the total weighted response time. Experiments are delicately designed on the testing instances with various settings. SAOF is compared with the baseline algorithm. Experimental results indicate that SAOF can more effectively reduce the total weighted response time of all tasks. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Heuristic algorithms

来源：评论

学校读者我要写书评

暂无评论

DRS: A deep reinforcement learning enhanced Kubernetes scheduler for microservice-based system

DRS: A deep reinforcement learning enhanced Kubernetes sched...

引用

作者： Jian, Zhaolong Xie, Xueshuo Fang, Yaozheng Jiang, Yibing Lu, Ye Dash, Ankan Li, Tao Wang, Guiling College of Computer Science Nankai University Tianjin China State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China College of Cyber Science Nankai University Tianjin China Department of Computer Science New Jersey Institute of Technology NewarkNJ United States

Recently, Kubernetes is widely used to manage and schedule the resources of microservices in cloud-native distributed applications, as the most famous container orchestration framework. However, Kubernetes preferentially schedules microservices to nodes with rich and balanced CPU and memory resources on a single node. The native scheduler of Kubernetes, called Kube-scheduler, may cause resource fragmentation and decrease resource utilization. In this paper, we propose a deep reinforcement learning enhanced Kubernetes scheduler named DRS. We initially frame the Kubernetes scheduling problem as a Markov decision process with intricately designed state, action, and reward structures in an effort to increase resource usage and decrease load imbalance. Then, we design and implement DRS mointor to perceive six parameters concerning resource utilization and create a thorough picture of all available resources globally. Finally, DRS can automatically learn the scheduling policy through interaction with the Kubernetes cluster, without relying on expert knowledge about workload and cluster status. We implement a prototype of DRS in a Kubernetes cluster with five nodes and evaluate its performance. Experimental results highlight that DRS overcomes the shortcomings of Kube-scheduler and achieves the expected scheduling target with three workloads. With only 3.27% CPU overhead and 0.648% communication delay, DRS outperforms Kube-scheduler by 27.29% in terms of resource utilization and reduces load imbalance by 2.90 times on average. © 2023 John Wiley & Sons, Ltd.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Formalization of Android Activity-Fragment Multitasking Mechanism and Static Analysis of Mobile Apps

引用

Formal Aspects of computing 2025年第2期37卷 1-86页

作者： He, Jinlong Wu, Zhilin Chen, Taolue Key Laboratory of System Software Chinese Academy of Sciences Institute of Software Chinese Academy of Sciences Beijing China State Key Laboratory of Computer Science Institute of Software Chinese Academy of Sciences Beijing China School of Computer Science and Technology University of Chinese Academy of Sciences Beijing China School of Computing and Mathematical Sciences Birkbeck University of London Northern Ireland London United Kingdom

The multitasking mechanism between activities and fragments plays a fundamental role in the Android operating system, which involves a wide range of features, including launch modes, intent flags, task affinities, and structured activities containing fragments. All of them are being widely used in Android apps, both open source and commercial ones. In this article, we present a formal semantics of the Android multitasking mechanism between activities and fragments, which accommodates all the important features and gives insofar the most comprehensive and accurate formalization. In particular, our semantics is formulated based on multi-stack systems, and fully captures the behavior of task stacks and activity stacks regarding fragments. Based on the semantics, we provide new static analysis algorithms, which are both multi-stack aware and fragment sensitive, thus achieve more precise static analysis for Android apps. We validate our approach by extensive experiments on both open source and commercial Android apps. The results highlight the benefits of the considering the semantics of the multitasking mechanism between activities and fragments in static analysis, and confirm the efficacy of our approach. © 2025 Copyright held by the owner/author(s).

关键词： Android multitasking mechanism activities and fragments formal semantics stack unboundedness static analysis

来源：评论

学校读者我要写书评

暂无评论

Block-gram: Mining Knowledgeable Features for Smart Contract Vulnerability Detection 7th

Block-gram: Mining Knowledgeable Features for Smart Contrac...

引用

7th International Conference on Smart computing and Communication, SmartCom 2022

作者： Li, Tao Wang, Haolong Fang, Yaozheng Jian, Zhaolong Wang, Zichun Xie, Xueshuo Tianjin Key Laboratory of Network and Data Security Technology Tianjin China College of Computer Science Nankai University Tianjin China Key Laboratory of Blockchain and Cyberspace Governance of Zhejiang Province Hangzhou China State Key Laboratory of Computer Architecture Institute of Computing Technology Chinese Academy of Sciences Beijing China

ISBN: (纸本)9783031281235

Effective vulnerability detection of large-scale smart contracts is critical because smart contract attacks frequently bring about tremendous economic loss. However, code analysis requiring traversal paths and learning methods requiring many features training is too time-consuming to detect large-scale on-chain contracts. This paper focuses on improving detection efficiency by reducing the dimension of the features, combined with expert knowledge. We propose a feature extraction method Block-gram to form low-dimensional knowledgeable features from the bytecode. We first separate the metadata and convert the runtime code to opcode sequence, dividing the opcode sequence into segments according to some instructions (jump, etc.). Then, we mine extensible Block-gram features for learning-based model training, consisting of 4-dimensional block features and 8-dimensional attribute features. We evaluate these knowledge-based features using seven state-of-the-art learning algorithms to show that the average detection latency speeds up 25 to 650 times, compared with the features extracted by N-gram. © 2023, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Smart contract

来源：评论

学校读者我要写书评

暂无评论

Selective Temporal Knowledge Graph Reasoning 30

Selective Temporal Knowledge Graph Reasoning

引用

Joint 30th International Conference on Computational Linguistics and 14th International Conference on Language Resources and Evaluation, LREC-COLING 2024

作者： Hou, Zhongni Jin, Xiaolong Li, Zixuan Bai, Long Guo, Jiafeng Cheng, Xueqi School of Computer Science and Technology University of Chinese Academy of Sciences China Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences China

ISBN: (纸本)9782493814104

Temporal Knowledge Graph (TKG), which characterizes temporally evolving facts in the form of (subject, relation, object, timestamp), has attracted much attention recently. TKG reasoning aims to predict future facts based on given historical ones. However, existing TKG reasoning models are unable to abstain from predictions they are uncertain, which will inevitably bring risks in real-world applications. Thus, in this paper, we propose an abstention mechanism for TKG reasoning, which helps the existing models make selective, instead of indiscriminate, predictions. Specifically, we develop a confidence estimator, called Confidence Estimator with History (CEHis), to enable the existing TKG reasoning models to first estimate their confidence in making predictions, and then abstain from those with low confidence. To do so, CEHis takes two kinds of information into consideration, namely, the certainty of the current prediction and the accuracy of historical predictions. Experiments with representative TKG reasoning models on two benchmark datasets demonstrate the effectiveness of the proposed CEHis. © 2024 ELRA Language Resource Association: CC BY-NC 4.0.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

HiRes: Hierarchical Feature Optimization and Rescorer for Automatic ICD Coding

HiRes: Hierarchical Feature Optimization and Rescorer for Au...

引用

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

作者： Liang, Zhenpeng Guan, Hongjiao Lu, Wenpeng Peng, Xueping Xu, Bing Yang, Muyun Key Laboratory of Computing Power Network and Information Security Ministry of Education Shandong Computer Science Center National Supercomputer Center in Jinan Qilu University of Technology Shandong Academy of Sciences Jinan China Shandong Provincial Key Laboratory of Computing Power Internet and Service Computing Shandong Fundamental Research Center for Computer Science Jinan China Australian Artificial Intelligence Institute University of Technology Sydney Sydney Australia Faculty of Computing Harbin Institute of Technology Harbin China

ISBN: (纸本)9798350368741

The International Classification of Diseases (ICD) coding assigns standardized codes to diseases. Automating this process enhances the efficiency and accuracy of clinical records processing. However, current methods struggle with noisy and lengthy clinical texts, making it difficult to ensure the reliability of feature extraction. Furthermore, they typically make separate binary predictions for each code, overlooking the dependencies between them. To address these issues, we propose a novel model called Hierarchical Feature Optimization and Rescorer (HiRes). We employ a cascaded convolution architecture to mitigate noise and enhance feature representation. Additionally, a masked autoencoder-based rescorer is introduced to capture ICD code interdependencies, refining initial predictions for improved accuracy. The model also considers the inconsistencies in code representations. Experiments on the MIMIC datasets demonstrate the effectiveness of our model. © 2025 IEEE.

关键词： automatic ICD coding cascaded convolutional network multi-label classification

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：