Aging-Aware Critical Path Selection via Graph Attention Networks

The graph attention network (GAT) [1] has started to become a mainstream neural network architecture since 2018, yielding remarkable performance gains in various natural language processing (NLP) tasks. Although GAT has reached the state-of-the-art (SOTA) performance as a recent success in relation prediction in knowledge graph, the current model is still limited by the following two aspects: (1) the existing model only considers the neighbors from the inbound-direction of the given entity, but ignores the rich neighborhood information from outbound-directions; (2) the existing model only uses the $k$ -th hop output to learn the multi-hop embeddings, which leads to the loss of a large amount of early-stage embedding information (e.g., one-hop) at the graph attention step. In this study, we propose a novel bidirectional graph attention network (BiGAT) to learn the hierarchical neighbor propagation. In our proposed BiGAT, an inbound-directional GAT and an outbound-directional GAT are introduced to capture sufficient neighborhood information before propagating the bidirectional neighborhood information to learn the multi-hop feature embeddings in a hierarchical manner. Experiments conducted on the four publicly available datasets show that BiGAT achieves the competitive results in comparison to other SOTA methods.

Graph Attention Network (GAT) has gained widespread adoption thanks to its exceptional performance in processing non-Euclidean graphs. The critical components of a GAT model involve aggregation and attention, which cause numerous main-memory access, occupying significant inference time. Recently, much research has proposed near-memory processing (NMP) architectures to accelerate aggregation. However, graph attention requires additional operations distinct from aggregation, making previous NMP architectures less suitable for supporting GAT, as they typically target aggregation-only workloads. In this paper, we propose GATe, a practical and efficient <underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">GAT</u> acc<underline xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">e</u>lerator with NMP architecture. To the best of our knowledge, this is the first time that accelerates both attention and aggregation computation on DIMM. We unify feature vector access to eliminate the two repetitive memory accesses to source nodes caused by the sequential phase-by-phase execution of attention and aggregation. Next, we refine the computation flow to reduce data dependencies in concatenation and softmax, which lowers on-chip memory usage and communication overhead. Additionally, we introduce a novel sharding method that enhances data reusability of high-degree nodes. Experiments show that GATe achieves substantial speedup of GAT attention and aggregation phases up to 6.77× and 2.46×, with average to 3.69× and 2.24×, respectively, compared to state-of-the-art NMP works GNNear and GraNDe.

For the first time, positive and negative bias temperature instability (P/NBTI) mechanisms in sub-nanometer EOT devices are investigated in this study. It is shown that PBTI degradation in sub-nanometer EOT devices occurs by interface degradation, additionally to the oxide bulk trap filling which is the dominant mechanism in over 1nm EOT devices. For NBTI, interface degradation remains as the main mechanism in sub-nano EOT devices, and additional high contribution of the high-k bulk defects can increase the degradation below 6A EOT.

The path decision problem for a hypersonic vehicle is recently formulated as it helps to decide a good initial guess for trajectory optimization with no-fly zone constraints. This path decision problem is a hybrid problem, discrete variables optimize paths and continuous variables optimize dynamics, which consumes costly computation time and is hard to apply in online scenarios. To reduce the computation, instead of directly solving the hybrid problem, we design a heuristics approach taking advantage of the interpretability and flexibility of a graph attention network (GAT). The path decision is modeled by a directed graph and transformed into a GAT training problem, and the resulting GAT can directly output a path in online use. During this work, there are two innovations: GAT customization and offline training. First, we define the mask to express the graph structure, and model the problem-specific decoder process in the GAT, thus meeting the path decision logic of no-fly zones avoidance and ensuring the solutions are feasible. Second, we numerically integrate dynamics by a path-following guidance law, calculate the total control effort as the cost function, and use this cost to train the GAT based on the widely used REINFORCE, thus conforming to dynamics that are practical for hypersonic vehicles. Simulation results illustrate the high-accuracy path decision, much faster calculation, and generalization on no-fly zone layouts and numbers.

With the introduction of High-k, metal gates and alternate substrates into the gate-stack at the 45nm and 32nm technology nodes, Bias Temperature Instability (BTI) phenomena have had to be included into the chip design modeling. In this paper, we explore BTI trends with High-k transistors in manufacturing ready CMOS processes with gate last and gate first type process flows. In both flows, Positive Bias Temperature Instability (PBTI) is a strong function of the interface and High-k thickness, with aggressive interface scaling having significant adverse reliability implications. Negative Bias Temperature Instability, on the other hand, is strongly dependent on the quality of the interface and its nitrogen content. The introduction of germanium into the Si channel is found to significantly improve NBTI. With recovery effects being strong in both NBTI and PBTI, AC BTI models in realistic circuit designs are critical to accurately evaluate the BTI lifetime of chips.

Identifying drug-target interactions (DTIs) is an important step in the process of new drug discovery and drug repositioning. Accurate predictions for DTIs can improve the efficiency in the drug discovery and development. Although rapid advances in deep learning technologies have generated various computational methods, it is still appealing to further investigate how to design efficient networks for predicting DTIs. In this study, we propose an end-to-end deep learning method (called MHSADTI) to predict DTIs based on the graph attention network and multi-head self-attention mechanism. First, the characteristics of drugs and proteins are extracted by the graph attention network and multi-head self-attention mechanism, respectively. Then, the attention scores are used to consider which amino acid subsequence in a protein is more important for the drug to predict its interactions. Finally, we predict DTIs by a fully connected layer after obtaining the feature vectors of drugs and proteins. MHSADTI takes advantage of self-attention mechanism for obtaining long-dependent contextual relationship in amino acid sequences and predicting DTI interpretability. More effective molecular characteristics are also obtained by the attention mechanism in graph attention networks. Multiple cross validation experiments are adopted to assess the performance of our MHSADTI. The experiments on four datasets, human, C.elegans, DUD-E and DrugBank show our method outperforms the state-of-the-art methods in terms of AUC, Precision, Recall, AUPR and F1-score. In addition, the case studies further demonstrate that our method can provide effective visualizations to interpret the prediction results from biological insights.

Due to the massive number of points and irregular structure, there is challenging to determine local graph relationships in large-scale point clouds. To address this problem, we propose a novel graph attention network that can embed global factors into neighboring points in specific local graphs. In this network, we introduce general local graph, which integrates all points sharing same order positions with encoded central point weights. In other words, it represents the most fundamental features of total local graphs. Besides, local graphs are regenerated to embed global factors by combining general local graph and central point attention. As a result, one point can obtain additional features from corresponding ones with same order positions in different neighborhoods. Experiment on large-scale point cloud segmentation dataset proves our network's competitive performance.

Aspect-based sentiment analysis is a fine-grained sentiment analysis that focuses on the sentiment polarity of different aspects of text, and most current research methods use a combination of dependent syntactic analysis and graphical neural networks. In this paper, a graph attention network aspect-based sentiment analysis model based on the weighting of dependencies (WGAT) is designed to address the problem in that traditional models do not sufficiently analyse the types of syntactic dependencies; in the proposed model, graph attention networks can be weighted and averaged according to the importance of different nodes when aggregating information. The model first transforms the input text into a low-dimensional word vector through pretraining, while generating a dependency syntax graph by analysing the dependency syntax of the input text and constructing a dependency weighted adjacency matrix according to the importance of different dependencies in the graph. The word vector and the dependency weighted adjacency matrix are then fed into a graph attention network for feature extraction, and sentiment polarity is predicted through the classification layer. The model can focus on syntactic dependencies that are more important for sentiment classification during training, and the results of the comparison experiments on the Semeval-2014 laptop and restaurant datasets and the ACL-14 Twitter social comment dataset show that the WGAT model has significantly improved accuracy and F1 values compared to other baseline models, validating its effectiveness in aspect-level sentiment analysis tasks.

With the booming logistics industry, multi-objective logistics transportation accounting for more than 60% and the annual logistics transportation volume growth rate of about 12%, traditional logistics path optimization algorithms are difficult to cope with complex and dynamic actual logistics scenarios. This paper proposes a multi-objective logistics path optimization model based on graph attention network and realizes real-time scheduling. By constructing the logistics network as a graph structure, using the graph attention mechanism to capture the complex relationship between nodes, designing innovative multi-objective fusion strategies and real-time information embedding mechanisms, the model is iteratively updated and its adaptability is enhanced. Experimental results show that in small-scale logistics networks, compared with traditional algorithms such as genetic algorithms, this model can reduce transportation costs by more than 20%; shorten transportation time by more than 25%; have the lowest cargo loss rate; and have the highest customer satisfaction. In scenarios such as large-scale logistics networks and complex road conditions, it also shows significant advantages, and the real-time scheduling response time is greatly shortened. For example, in the dynamic traffic change scenario in Case 1, the path adjustment can be completed in only 5 min, which is much faster than other comparison models. This study provides efficient operation solutions for logistics companies and promotes the development of the logistics industry towards intelligence and efficiency.

Recommendation systems is the core of most online experiences, guiding people to the right products, films, music, courses, and even social links when the choices are overwhelming. They work by learning from large interaction logs and applying predictive models to surface content that fits each user’s tastes, steadily improving as algorithms and data grow richer. As AI and deep learning have developed, classic strategies like collaborative filtering and content-based filtering are increasingly paired with, or superseded by, neural architectures that can model more nuanced patterns in behaviour and content. Among these neural approaches, graph neural networks—and especially Graph Attention Networks (GATs)—have become influential for recommendation because they treat users and items as nodes in a graph and learn from the connections between them. With attention mechanisms, GATs decide which neighbours matter most for a given prediction, amplifying useful signals and dampening noise to learn stronger user and item representations. This survey traces how the industry moved from traditional methods to graph-based models centred on GATs, comparing approaches on scalability, ranking accuracy, data sparsity, and interpretability across widely used benchmarks. Drawing from recent peer‑reviewed work, it highlights architectures and training practices that consistently help in real-world settings, as well as gaps that still limit robust deployment at scale. Looking forward, several directions stand out: using explainable AI to clarify why specific items are recommended, combining reinforcement learning with GATs for adaptive policies, and adopting federated learning to protect user privacy while training on distributed data. The versatility of GAT-based recommenders is already visible across domains—from e‑commerce and online learning to healthcare and social feeds—where graph structure and selective attention naturally fit the data. Overall, the evidence points to GATs as a key ingredient for the next generation of personalized, context‑aware, and trustworthy recommendation systems. Keywords: Recommendation Systems, Collaborative Filtering, Graph Neural Networks, Graph Attention Networks

Circular RNAs (circRNAs) are covalently closed single-stranded RNA molecules, which have many biological functions. Previous experiments have shown that circRNAs are involved in numerous biological processes, especially regulatory functions. It has also been found that circRNAs are associated with complex diseases of human beings. Therefore, predicting the associations of circRNA with disease (called circRNA-disease associations) is useful for disease prevention, diagnosis and treatment. In this work, we propose a novel computational approach called GGCDA based on the Graph Attention Network (GAT) and Graph Convolutional Network (GCN) to predict circRNA-disease associations. Firstly, GGCDA combines circRNA sequence similarity, disease semantic similarity and corresponding Gaussian interaction profile kernel similarity, and then a random walk with restart algorithm (RWR) is used to obtain the preliminary features of circRNA and disease. Secondly, a heterogeneous graph is constructed from the known circRNA-disease association network and the calculated similarity of circRNAs and diseases. Thirdly, the multi-head Graph Attention Network (GAT) is adopted to obtain different weights of circRNA and disease features, and then GCN is employed to aggregate the features of adjacent nodes in the network and the features of the nodes themselves, so as to obtain multi-view circRNA and disease features. Finally, we combined a multi-layer fully connected neural network to predict the associations of circRNAs with diseases. In comparison with state-of-the-art methods, GGCDA can achieve AUC values of 0.9625 and 0.9485 under the results of fivefold cross-validation on two datasets, and AUC of 0.8227 on the independent test set. Case studies further demonstrate that our approach is promising for discovering potential circRNA-disease associations.

The explicit knowledge contained in quality acceptance specifications is crucial for the quality management of infrastructure projects. As the digitalization of knowledge in the architecture, engineering, and construction industry accelerates, the need for automated knowledge mining to process large volumes of regulatory documents is increasing. However, challenges such as complex relationships, entity overlap, and long-distance entities in these specifications pose significant difficulties for knowledge extraction. In this research, an autonomous knowledge mining framework based on the cascade binary tagging framework (CasRel) and graph attention network (GAT) is proposed, which is designed to transform complex textual knowledge from quality acceptance specifications into graph-based knowledge representations. The framework utilizes the RoBERTa-wwm-ext layer to encode the input text, employs a subject tagger to extract entity start and end positions, and constructs a graph structure. Using the graph attention mechanism in GAT, features of neighboring nodes are aggregated to generate context-enhanced entity feature representations. A relation-specific object tagger then outputs complete knowledge triples. A Neo4j database is used to store and visualize the extracted knowledge triples, creating a knowledge graph with 3,762 nodes. Using the Chinese railway construction quality acceptance specifications as a test case, the results demonstrate that the suggested framework achieves excellent performance in sophisticated knowledge extraction and knowledge graph modeling. This research effectively transforms textual regulatory documents into structured knowledge bases, supporting automated knowledge acquisition and the digital management of compliance information.

Graph Neural Networks (GNNs) are deep learning models specifically designed for analyzing graph-structured data, capturing complex relationships and structures to improve analysis and prediction. A common task in GNNs is node classification, where each node in the graph is assigned a predefined category. The Graph Attention Network (GAT) is a popular variant of GNNs known for its ability to capture complex dependencies by assigning importance weights to nodes during information aggregation. However, the GAT’s reliance on local attention mechanisms limits its effectiveness in capturing global information and long-range dependencies. To address this limitation, we propose a new attention mechanism called Global-Local Graph Attention (GLGA). Our mechanism enables the GAT to capture long-range dependencies and global graph structures while maintaining its ability to focus on local interactions. We evaluate our algorithm on three citation datasets (Cora, Citeseer, and Pubmed) using multiple metrics, demonstrating its superiority over other baseline models. The proposed GLGA mechanism has been proven to be an effective solution for improving node classification tasks.

SiC-MOSFETs have been markedly developed, and the main issues remaining are high-temperature stability and gate-oxide integrity. In this paper, we report special features of Ids–Vgs and Ig–Vgs characteristics of commercially available SiC-MOSFETs during high-gate-voltage and high-temperature stress. Moreover, we introduce simple analytical models that reveal the effects of the trapped charges on the characteristics. Interface states around the conduction-band edge are suggested to be the key mechanisms of Fowler–Nordheim stress degradation (FN degradation) as well as negative bias temperature instability (NBTI), which affects both mobility and carrier density. Hole trapping is also important in FN degradation as well as NBTI. Near-interfacial electron trapping and detrapping found in positive bias temperature instability (PBTI) is also observed in FN degradation. A continuous Ig increase during FN stress, which is the most specific to SiC-MOSFETs, is related to the observation that near-interfacial electron trapping has little effect on FN tunneling.

Abstract. With introduction of high-k gate oxide materials, the degradation effect Positive Bias Temperature Instability (PBTI) is starting to play an important role. Together with the still effective Negative Bias Temperature Instability (NBTI) it has significant influence on the 6T SRAM memory cell. We present simulations of both effects, first isolated, then combined in SRAM operation. During long hold of one data, both effects add up to a worst case impact. This leads to an asymmetric cell, which, in a directly following read cycle, combined with the generally unavoidable production variations, maximizes the risk of destructive reading. In future SRAM designs, it will be important to consider this combination of effects to avoid an undesired write event.