A5103092072- Advanced Graph Neural Networks
- Parallel Computing and Optimization Techniques
- Advanced Neural Network Applications
- Recommender Systems and Techniques
- Liver Disease Diagnosis and Treatment
- Advanced Data Storage Technologies
- Rheumatoid Arthritis Research and Therapies
- Stochastic Gradient Optimization Techniques
- Tensor decomposition and applications
- Graph Theory and Algorithms
- Lexicography and Language Studies
- Fatty Acid Research and Health
- Advanced Computational Techniques and Applications
- Service-Oriented Architecture and Web Services
- Complex Network Analysis Techniques
- Bone Metabolism and Diseases
- Sesame and Sesamin Research
- Machine Learning in Bioinformatics
- Matrix Theory and Algorithms
- Technology and Security Systems
- Inflammatory Bowel Disease
- Mathematics, Computing, and Information Processing
- Solar Radiation and Photovoltaics
- Hepatitis B Virus Studies
- Neurological Disease Mechanisms and Treatments
Kunming University of Science and Technology
2024
Tsinghua University
2021-2023
University of California, San Diego
2023
Universidad Católica Santo Domingo
2023
Shaoxing People's Hospital
2009-2022
Liupanshui Normal University
2017
Zhejiang University
2014
Education Department of Jiangxi Province
2005-2007
Nanchang Normal University
2005-2007
Recent developments in Large Language Model (LLM) agents are revolutionizing Autonomous Software Engineering (ASE), enabling automated coding, problem fixes, and feature improvements. However, localization -- precisely identifying software problems by navigating to relevant code sections remains a significant challenge. Current approaches often yield suboptimal results due lack of effective integration between LLM precise search mechanisms. This paper introduces OrcaLoca, an agent framework...
Point cloud computation has become an increasingly more important workload for autonomous driving and other applications. Unlike dense 2D computation, point convolution sparse irregular patterns thus requires dedicated inference system support with specialized high-performance kernels. While existing deep learning libraries have developed different dataflows on clouds, they assume a single dataflow throughout the execution of entire model. In this work, we systematically analyze improve...
Flavonoids are the active component of Herba Epimedii (H. Epimedii), which is commonly used in Asia. This study to investigate effect H. on bone repair after anti-infection treatment vivo. The bioactive-composition group (BCGE) contained four flavonoids with total content 43.34%. Rabbits chronic osteomyelitis response injection Staphylococcus aureus were treated BCGE 242.70 mg/kg/day intragastrically vancomycin-calcium sulphate treatment. Micro-computerd tomography (CT), morphology, blood...
Graph neural networks (GNNs) have attracted tremendous attention from the graph learning community in recent years. It has been widely adopted various real-world applications diverse domains, such as social and biological graphs. The research of deep present new challenges, including sparse nature data, complicated training GNNs, non-standard evaluation tasks. To tackle issues, we CogDL1, a comprehensive library for that allows researchers practitioners to conduct experiments, compare...
Sparse convolution plays a pivotal role in emerging workloads, including point cloud processing AR/VR, autonomous driving, and graph understanding recommendation systems. Since the computation pattern is sparse irregular, specialized high-performance kernels are required. Existing GPU libraries offer two dataflow types for convolution. The gather-GEMM-scatter easy to implement but not optimal performance, while dataflows with overlapped memory access (e.g. implicit GEMM) highly performant...
Sparse Matrix-Matrix Multiplication (SpMM) has served as fundamental components in various domains. Many previous studies exploit GPUs for SpMM acceleration because provide high bandwidth and parallelism. We point out that a static design does not always improve the performance of on different input data (e.g., >85% loss with single algorithm). In this paper, we consider challenge dynamics from novel auto-tuning perspective, while following issues remain to be solved: (1) Orthogonal...
Graph Neural Networks (GNNs) have been widely used in various domains, and GNNs with sophisticated computational graph lead to higher latency larger memory consumption. Optimizing the GNN suffers from: (1) Redundant neural operator computation. The same data are propagated through structure perform operation multiple times GNNs, leading redundant computation which accounts for 92.4% of total operators. (2) Inconsistent thread mapping. Efficient mapping schemes vertex-centric edge-centric...
Protein structure prediction helps to understand gene translation and protein function, which is of growing interest importance in structural biology. The AlphaFold model, used transformer architecture achieve atomic-level accuracy prediction, was a significant breakthrough. However, training inference the model are challenging due its high computation memory cost. In this work, we present FastFold, an efficient implementation for both inference. We propose Dynamic Axial Parallelism Duality...
microRNAs have been reported to play important roles in the pathogenesis of rheumatoid arthritis (RA). This study examined effects miR-522 on biological behaviors RA synovial fibroblasts. The expression levels and relevant genes were measured by quantitative real-time PCR. protein cytokines determined ELISA assay. matrix metalloproteinases (MMPs) suppressor cytokine signaling 3 (SOCS3) western blot Luciferase reporter assay was used confirm potential target miR-522. Our results showed that...
The coronavirus disease 2019 (COVID-19) pandemic has imposed enormous morbidity and mortality burdens. Patients with rheumatic diseases (RDs) are vulnerable to the COVID-19 infection, given their immunocompromised status. Ensuring acceptance of vaccine is important attracted attention by health professionals. In this study, we designed an online cross-sectional survey that used questionnaire from 8 May 2021 4 October 2021. Attitudes toward vaccination, personal information, current activity...
In response to the issue of short-term fluctuations in photovoltaic (PV) output due cloud movement, this paper proposes a method for forecasting PV based on Depthwise Separable Convolution Visual Geometry Group (DSCVGG) and Deep Gate Recurrent Neural Network (DGN). Initially, motion prediction model is constructed using DSCVGG, which achieves edge recognition clouds by replacing previous convolution layer pooling VGG with depthwise separable convolution. Subsequently, results DSCVGG network,...
The aim of this study was to evaluate disease-activity-guided stepwise tapering or discontinuation rhTNFR:Fc, an etanercept biosimilar, in patients with ankylosing spondylitis (AS) a prospective, randomized, open-label, multicentric study.Active AS disease activity score (ASDAS) ⩾2.1 recruited from 10 hospitals were treated rhTNFR:Fc 50 mg weekly for 12 weeks, and further randomized into different groups according ASDAS at week 12. Patients who achieved clinical remission (ASDAS < 1.3)...
Sampled Dense-Dense Matrix Multiplication (SDDMM) is a core component of many machine learning systems. SDDMM exposes substantial amount parallelism that favors throughput-oriented architectures like the GPU. However, accelerating it on GPUs challenging in two aspects: poor memory access locality caused by sparse sampling matrix with dot-product reduction vectors dense matrices. To address both challenges, we present PRedS to boost efficiency suite Parallel Reduction Scheduling...
Graph neural networks (GNNs) have attracted tremendous attention from the graph learning community in recent years. It has been widely adopted various real-world applications diverse domains, such as social and biological graphs. The research of deep present new challenges, including sparse nature data, complicated training GNNs, non-standard evaluation tasks. To tackle issues, we CogDL, a comprehensive library for that allows researchers practitioners to conduct experiments, compare...
The automatic generation of RTL code (e.g., Verilog) through natural language instructions has emerged as a promising direction with the advancement large models (LLMs). However, producing that is both syntactically and functionally correct remains significant challenge. Existing single-LLM-agent approaches face substantial limitations because they must navigate between various programming languages handle intricate generation, verification, modification tasks. To address these challenges,...
In recent years, Graph Neural Networks (GNNs) have ignited a surge of innovation, significantly enhancing the processing geometric data structures such as graphs, point clouds, and meshes. As domain continues to evolve, series frameworks libraries are being developed push GNN efficiency new heights. While graph-centric achieved success in past, advent efficient tensor compilers has highlighted urgent need for tensor-centric libraries. Yet, GNNs remain scarce due unique challenges limitations...
This paper presents a new method of liquid surface location based on visual analysis, which was proposed with the navigation imagery processing technology. Firstly, hospital medical infusion bottle image binarized by statistical methods energy. Secondly, binary carried segmentation through projection statistics and Shen algorithm. Finally, horizontal is smoothed The experimental results show that can accurately locate real-time tracking surface.
Triangle counting (TC) is one of the most fundamental graph analysis tools with a wide range applications. Modern triangle algorithms traverse and perform set intersections neighbor sets to find triangles. However, existing approaches suffer from heavy off-chip memory access intersection overhead. Thus, we propose CLAP, first content addressable (CAM) based architecture software hardware co-optimizations. To reduce number intersections, force-based node index reorder method. It...