检索结果-内蒙古大学图书馆

Parenting: Optimizing Knowledge Selection of Retrieval-Augmented Language Models with Parameter Decoupling and Tailored Tuning

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Xu, Yongxin Zhang, Ruizhe Jiang, Xinke Feng, Yujie Xiao, Yuzhen Ma, Xinyu Zhu, Runchuan Chu, Xu Zhao, Junfeng Wang, Yasha School of Computer Science School of Software & Microelectronics Peking University Beijing China Key Laboratory of High Confidence Software Technologies Ministry of Education Beijing China Center on Frontiers of Computing Studies Peking University Beijing China National Engineering Research Center For Software Engineering Peking University Beijing China Department of Computing The Hong Kong Polytechnic University Hong Kong

Retrieval-Augmented Generation (RAG) offers an effective solution to the issues faced by Large Language Models (LLMs) in hallucination generation and knowledge obsolescence by incorporating externally retrieved knowledge. However, due to potential conflicts between internal and external knowledge, as well as retrieval noise, LLMs often struggle to effectively integrate external evidence, leading to a decline in performance. Although existing methods attempt to tackle these challenges, they often struggle to strike a balance between model adherence and robustness, resulting in significant learning variance. Inspired by human cognitive processes, we propose Parenting, a novel framework that decouples adherence and robustness within the parameter space of LLMs. Specifically, Parenting utilizes a key parameter mining method based on forward activation gain to identify and isolate the crucial parameter units that are strongly linked to adherence and robustness. Then, Parenting employs a type-guided tailored tuning strategy, applying specific and appropriate fine-tuning methods to parameter units representing different capabilities, aiming to achieve a balanced enhancement of adherence and robustness. Extensive experiments on various datasets and models validate the effectiveness and generalizability of our methods. Copyright © 2024, The Authors. All rights reserved.

关键词： Obsolescence

Continuity Scaling: A Rigorous Framework for Detecting and Quantifying Causality Accurately

学校读者我要写书评

暂无评论

research 2022年第4期2022卷 137-146页

作者： Xiong Ying Si-Yang Leng Huan-Fei MaD Qing Nie Ying-Cheng Lai Wei Lin School of Mathematical Sciences SCMSand SCAMFudan UniversityShanghai 200433China Research Institute for Inteligent Complex Systems CCSBand LCNBIFudan UniversityShanghai 200433China State Key Laboratory of Medical Neurobiology and MOE Frontiers Center for Brain Science Institutes of Brain ScienceFudan UniversityShanghai 200032China Institute of AI and Robotics Academy for Engineering and TechnologyFudan UniversityShanghai 200433China School of Mathematical Sciences Soochow UniversitySuzhou 215006China Department of Mathematics Department of Developmental and Cll BiologyAnd NSF-Simons Center for Multiscale Cell Fate ResearchUniversity of CaliforniaIrvineCA 92697-3875USA School of Electrical ComputerAnd Energy EngineeringArizona State UniversityTempeArizona 85287-5706USA Shanghai Artificial Intelligence Laboratory Shanghai 200232China

Data-based detection and quantification of causation in complex,nonlinear dynamical systems is of paramount importance to science,engineering,and *** by the widely used methodology in recent years,the cross-map-based techniques,we develop a general framework to advance towards a comprehensive understanding of dynamical causal mechanisms,which is consistent with the natural interpretation of *** particular,instead of measuring the smoothness of the cross-map as conventionally implemented,we define causation through measuring the scaling law for the continuity of the investigated dynamical system *** uncovered scaling law enables accurate,reliable,and eficient detection of causation and assessment of its strength in general complex dynamical systems,outperforming those existing representative *** continuity scaling-based framework is rigorously established and demonstrated using datasets from model complex systems and the real world.

关键词： dynamical scaling continuity

Advanced wind turbine blade inspection with hyperspectral imaging and 3D convolutional neural networks for damage detection

学校读者我要写书评

暂无评论

Energy and AI 2024年第2期16卷 528-537页

作者： Patrick Rizk Frederic Rizk Sasan Sattarpanah Karganroudi Adrian Ilinca Rafic Younes Jihan Khoder Department of Mechanical Engineering Universitédu QuébecàTrois-RivièresDrummondvilleQCCanada Centre national intégrédu manufacturier intelligent(CNIMI) Universitédu QuébecàTrois-RivièresDrummondvilleQCCanada Department of Mathematics Computer Science and EngineeringUniversitédu QuébecàRimouski300 allée des UrsulinesRimouskiQCG5L3A1Canada The Center for Advanced Computer Studies University of Louisiana at LafayetteLafayetteLAUSA T3E Industrial Research Group Mechanical Engineering DepartmentÉcole de Technologie Supérieure1100 Notre-Dame St WMontréalQC H3C 1K3Canada Faculty of Engineering Third BranchLebanese UniversityHadathBeirutLebanon LISV Laboratory University of Versailles Saint-Quentin-en-Yvelines10-12 Avenue de l'Europe78140VélizyFrance

In the context of global efforts to mitigate climate change by pursuing sustainable energy sources, wind energy has emerged as a critical contributor. However, the wind energy industry faces substantial challenges in maintaining and preserving the integrity of wind turbine blades. Timely and accurate detection and classification of blade faults, encompassing issues such as cracks, erosion, and ice buildup, are imperative to uphold wind turbines' ongoing efficiency and safety. This study introduces an inventive approach that amalgamates hyperspectral imaging and 3D Convolutional Neural Networks (CNNs) to augment the precision and efficiency of wind turbine blade fault detection and classification. Hyperspectral imaging is harnessed to capture comprehensive spectral information from blade surfaces, facilitating exact fault identification. The process is streamlined through Incremental Principal Component Analysis (IPCA), reducing data dimensions while maintaining integrity. The 3D CNN model demonstrates remarkable performance, achieving high accuracy in detecting all fault categories in full-band hyperspectral images. The model retains high accuracy even with dimensionality reduction to 20 spectral bands. The reduced processing time of the 20-band image enhances the practicality of real-world applications, thereby reducing downtime and maintenance expenditures. This research represents a significant advancement in wind turbine blade inspection, contributing to the sustainability and dependability of wind energy systems and furthering the cause of a cleaner and more sustainable energy future as part of the broader fight against climate change.

关键词： blade thereby turbine

A Corrected Inexact Proximal Augmented Lagrangian Method with a Relative Error Criterion for a Class of Group-quadratic Regularized Optimal Transport Problems

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Yang, Lei Liang, Ling Chu, Hong T.M. Toh, Kim-Chuan School of Computer Science and Engineering Guangdong Province Key Laboratory of Computational Science Sun Yat-Sen University China Department of Mathematics University of Maryland College Park United States Department of Mathematics National University of Singapore Singapore Department of Mathematics Institute of Operations Research and Analytics National University of Singapore Singapore119076 Singapore

The optimal transport (OT) problem and its related problems have attracted significant attention and have been extensively studied in various applications. In this paper, we focus on a class of group-quadratic regularized OT problems which aim to find solutions with specialized structures that are advantageous in practical scenarios. To solve this class of problems, we propose a corrected inexact proximal augmented Lagrangian method (ciPALM), with the subproblems being solved by the semi-smooth Newton (Ssn) method. We establish that the proposed method exhibits appealing convergence properties under mild conditions. Moreover, our ciPALM distinguishes itself from the recently developed semismooth Newton-based inexact proximal augmented Lagrangian (Snipal) method for linear programming. Specifically, Snipal uses an absolute error criterion for the approximate minimization of the subproblem for which a summable sequence of tolerance parameters needs to be pre-specified for practical implementations. In contrast, our ciPALM adopts a relative error criterion with a single tolerance parameter, which would be more friendly to tune from computational and implementation perspectives. These favorable properties position our ciPALM as a promising candidate for tackling large-scale problems. Various numerical studies validate the effectiveness of employing a relative error criterion for the inexact proximal augmented Lagrangian method, and also demonstrate that our ciPALM is competitive for solving large-scale group-quadratic regularized OT *** Codes 90C05, 90C06, 90C25 Copyright © 2023, The Authors. All rights reserved.

关键词： Linear programming

Physiological accuracy in simulating refractory cardiac tissue: the volume-averaged bidomain model vs. the cell-based EMI model

学校读者我要写书评

暂无评论

arXiv 2023年

作者： Reimer, Joyce Domínguez-Rivera, Sebastián A. Sundnes, Joakim Spiteri, Raymond J. Division of Biomedical Engineering University of Saskatchewan Saskatoon Canada Department of Mathematics and Statistics University of Saskatchewan Saskatoon Canada Simula Research Laboratory Oslo Norway Department of Computer Science University of Saskatchewan Saskatoon Canada

The refractory period of cardiac tissue can be quantitatively described using strength-interval (SI) curves. The information captured in SI curves is pertinent to the design of anti-arrhythmic devices including pacemakers and implantable cardioverter defibrillators. As computational cardiac modelling becomes more prevalent, it is feasible to consider the generation of computationally derived SI curves as a supplement or precursor to curves that are experimentally derived. It is beneficial, therefore, to examine the profiles of the SI curves produced by different cardiac tissue models to determine whether some models capture the refractory period more accurately than others. In this study, we compare the unipolar SI curves of two tissue models: the current state-of-the-art bidomain model and the recently developed extracellular-membrane-intracellular (EMI) model. The EMI model’s resolution of individual cell structure makes it a more detailed model than the bidomain model, which forgoes the structure of individual cardiac cells in favour of treating them homogeneously as a continuum. We find that the resulting SI curves elucidate differences between the models, including that the behaviour of the EMI model is noticeably closer to the refractory behaviour of experimental data compared to that of the bidomain model. These results hold implications for future computational pacemaker simulations and shed light on the predicted refractory properties of cardiac tissue from each model. © 2023, CC BY.

关键词： Tissue

Enhancing the Sum Rate in Massive MIMO Optical Wireless Communication System by Using MMSE Precoding

学校读者我要写书评

暂无评论

Enhancing the Sum Rate in Massive MIMO Optical Wireless Comm...

International Conference on Communication, Computing and Digital Systems (C-CODE)

作者： Zakir Ali Muhammad Furqan Qasim Jan National Mobile Communications Research Laboratory Southeast University Nanjing China Department of Computer Sciences Bahria University Lahore Campus Pakistan Department of Software Engineering Capital University of Science and Technology (CUST) Islamabad Pakistan Department of Computer Science COMSATS University Islamabad Attock Campus Attock Pakistan

We investigate an optical wireless communications system based on massive multiple-input multiple-output (MIMO). Through a transmit lens, the optical base station with its huge optical transmitters communicates with the user terminals (UTs). We create a channel model for optical massive MIMO transmissions and examine light refraction caused by lenses, with an emphasis on LED transmitters. We formulate the sum rate maximization problem and convert it into a convex optimization problem to get a closed-form solution. MMSE is used to increase the sum rate and Karush-Kuhn-Tucker (KKT) conditions were used to solve the aforementioned problem. We study traditional precoding techniques, maximum ratio transmission (MRT), and regularized zero-forcing precoding (RZF) for an optical massive MIMO system. The proposed technique is ajoint precoder design to enhance the sum rate. The simulation results show that the proposed MMSE precoder design outperforms MRT and RZF precoders in a massive MIMO optical wireless communication system.

关键词：

Nonparametric Sparse Online Learning of the Koopman Operator

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Hou, Boya Sanjari, Sina Dahlin, Nathan Koppel, Alec Bose, Subhonmesh Carl R. Woese Institute for Genomic Biology University of Illinois Urbana-Champaign United States Department of Mathematics and Computer Science Royal Military College of Canada Canada Department of Electrical and Computer Engineering University at Albany SUNY United States Artificial Intelligence Research JP Morgan Chase & Co United States Department of Electrical and Computer Engineering Coordinated Science Laboratory University of Illinois Urbana-Champaign United States

The Koopman operator provides a powerful framework for representing the dynamics of general nonlinear dynamical systems. Data-driven techniques to learn the Koopman operator typically assume that the chosen function space is closed under system dynamics. In this paper, we study the Koopman operator via its action on the reproducing kernel Hilbert space (RKHS), and explore the mis-specified scenario where the dynamics may escape the chosen function space. We relate the Koopman operator to the conditional mean embeddings (CME) operator and then present an operator stochastic approximation algorithm to learn the Koopman operator iteratively with control over the complexity of the representation. We provide both asymptotic and finite-time last-iterate guarantees of the online sparse learning algorithm with trajectory-based sampling with an analysis that is substantially more involved than that for finite-dimensional stochastic approximation. Numerical examples confirm the effectiveness of the proposed algorithm. © 2024, CC BY-NC-ND.

关键词： Dynamical systems

AircraftVerse: a large-scale multimodal dataset of aerial vehicle designs 23

学校读者我要写书评

暂无评论

AircraftVerse: a large-scale multimodal dataset of aerial ve...

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Adam D. Cobb Anirban Roy Daniel Elenius F. Michael Heim Brian Swenson Sydney Whittington James D. Walker Theodore Bapty Joseph Hite Karthik Ramani Christopher McComb Susmit Jha Computer Science Laboratory SRI International Mechanical Engineering Division Southwest Research Institute Intelligent Systems Division Southwest Research Institute Institute of Software Integrated Systems Vanderbilt University School of Mechanical Engineering Purdue University Department of Mechanical Engineering Carnegie Mellon University

We present AircraftVerse, a publicly available aerial vehicle design dataset. Aircraft design encompasses different physics domains and, hence, multiple modalities of representation. The evaluation of these cyber-physical system (CPS) designs requires the use of scientific analytical and simulation models ranging from computer- aided design tools for structural and manufacturing analysis, computational fluid dynamics tools for drag and lift computation, battery models for energy estimation, and simulation models for flight control and dynamics. AircraftVerse contains 27,714 diverse air vehicle designs - the largest corpus of engineering designs with this level of complexity. Each design comprises the following artifacts: a symbolic design tree describing topology, propulsion subsystem, battery subsystem, and other design details; a STandard for the Exchange of Product (STEP) model data; a 3D CAD design using a stereolithography (STL) file format; a 3D point cloud for the shape of the design; and evaluation results from high fidelity state-of-the-art physics models that characterize performance metrics such as maximum flight distance and hover-time. We also present baseline surrogate models that use different modalities of design representation to predict design performance metrics, which we provide as part of our dataset release. Finally, we discuss the potential impact of this dataset on the use of learning in aircraft design and, more generally, in CPS. AircraftVerse is accompanied by a data card, and it is released under Creative Commons Attribution-ShareAlike (CC BY-SA) license. The dataset is hosted at https://***/record/6525446 baseline models and code at https://***/SRI-CSL/AircraftVerse and the dataset description at https://***/.

关键词：

Invertible Residual Rescaling Models

学校读者我要写书评

暂无评论

arXiv 2024年

作者： Li, Jinmin Dai, Tao Zha, Yaohua Luo, Yilu Lu, Longfei Chen, Bin Wang, Zhi Xia, Shu-Tao Zhang, Jingyun Tsinghua Shenzhen International Graduate School Tsinghua University China College of Computer Science and Software Engineering Shenzhen University China Research Center of Artificial Intelligence Peng Cheng Laboratory China Department of Computer Science and Technology Harbin Institute of Technology Shenzhen China WeChat Pay Lab33 Tencent China

Invertible Rescaling Networks (IRNs) and their variants have witnessed remarkable achievements in various image processing tasks like image rescaling. However, we observe that IRNs with deeper networks are difficult to train, thus hindering the representational ability of IRNs. To address this issue, we propose Invertible Residual Rescaling Models (IRRM) for image rescaling by learning a bijection between a high-resolution image and its low-resolution counterpart with a specific distribution. Specifically, we propose IRRM to build a deep network, which contains several Residual Downscaling Modules (RDMs) with long skip connections. Each RDM consists of several Invertible Residual Blocks (IRBs) with short connections. In this way, RDM allows rich low-frequency information to be bypassed by skip connections and forces models to focus on extracting high-frequency information from the image. Extensive experiments show that our IRRM performs significantly better than other state-of-the-art methods with much fewer parameters and complexity. Particularly, our IRRM has respectively PSNR gains of at least 0.3 dB over HCFlow and IRN in the ×4 rescaling while only using 60% parameters and 50% FLOPs. The code will be available at https://***/THUKingmin/IRRM. Copyright © 2024, The Authors. All rights reserved.

关键词： Machine learning