Graph neural networks (GNNs) have demonstrated strong performance in various graph mining tasks but rely heavily on extensively labeled nodes. To improve training efficiency, graph active learning (GAL) has emerged as a solution for selecting the most informative nodes for labeling. However, existing GAL methods are primarily designed for homophilic graphs, where nodes with the same labels are more likely to be connected. In this work, we systematically study active learning on heterophilic graphs, a setting that has received limited attention. Surprisingly, we observe that existing GAL methods fail to consistently outperform random sampling on heterophilic graphs. Through an in-depth investigation, we reveal that these methods implicitly assume homophily even on heterophilic graphs, leading to suboptimal performance. To address this issue, we introduce the principle of "Know Your Neighbors" and propose an active learning algorithm KyN specifically for heterophilic graphs. The core idea of KyN is to provide GNNs with accurate estimations of homophily distribution by labeling nodes together with their neighbors. We implement KyN based on subgraph sampling with probabilities proportional to l1 Lewis weights, which is supported by solid theoretical guarantees. Extensive experiments on diverse real-world datasets, including a large heterophilic graph with over 2 million nodes, demonstrate the effectiveness and scalability of KyN.

Source Address Validation (SAV) is a security mechanism deployed at network boundaries to prevent packets with illegal source addresses from crossing these boundaries. While SAV plays an important role in mitigating source address spoofing, its deployment across the global Internet remains limited. Measuring SAV deployment is essential for enhancing the understanding of the security landscape of networks. In this study, we propose a novel method for measuring SAV deployment based on 6 in 4 tunnels, complementing existing measurement work. Using this method, we measure inbound SAV for IPv4 and outbound SAV for IPv6, obtaining results from 12,417 and 2,104 Autonomous Systems (ASes), respectively. Based on our measurements, we analyze factors that may influence SAV deployment, including network address space size, AS type, and geographical location. Additionally, we provide a global heatmap of spoofable rates for networks in different countries and regions. Note that the measuring method using bin4 tunnels we introduce is not only applicable to SAV measurements but also holds potential for other measurement tasks, such as connectivity testing and transmission path discovery. This method offers a new way for large-scale measurement tasks across different networks, which may benefit future research.

Graph recommender (GR) is a type of graph neural network (GNN) encoder that is customized for extracting information from the user-item interaction graph. Due to its strong performance on the recommendation task, GR has gained significant attention recently. Graph contrastive learning (GCL) is also a popular research direction that aims to learn, often unsupervised, GNNs with certain contrastive objectives. As general graph representation learning methods, GCLs have been widely adopted with supervised recommendation loss for joint training of GRs. Despite the intersection of GR and GCL research, theoretical understanding of the relationship between the two fields is surprisingly sparse. This vacancy inevitably leads to inefficient scientific research.

Intelligent computing center (ICC) is a new type of data center constructed with intelligent computing power, such as graphic processing units (GPUs) and artificial intelligence acceleration cards. With billions of parameters, the emergence of large models (e.g., ChatGPT) presents a significant demand of computing power. It may be challenging for a single ICC to provide the required computing power during large model training. Thus, ICC interconnections (ICCI) will become a typical and effective solution to provide intensive computing power. Due to human activities, traditional computing tasks (e.g., transaction processing and online entertainment) exhibit a tidal effect of computing demand, which leads to the tidal variation of remaining computing resources. Moreover, distributed model training (DMT) tasks are likely to cover peaks and valleys of the tidal effect in computing power. In this case, it is easy for DMT tasks to cause an ICC to become a hotspot (i.e., computing load in an ICC exceeds a desired threshold), which significantly degrades the reliability and performance of the ICC. This paper proposes DeepHM, a deep reinforcement learning-based hotspot management strategy through task migration in ICCI networks. To comprehensively consider the bandwidth metrics of the ICCI network, we further propose a dynamic wavelength allocation strategy, i.e., DeepHM-DWA. Simulation results show that the DeepHM and DeepHM-DWA reduce the hotspot compute unit time blocks by 19% and 18% with fewer number of migrated workers while balancing the computing load among multiple ICCs. DeepHM and DeepHM-DWA reduce the average completion time ratio of the DMT tasks by 2% and 5%, respectively.

As native IPv6 adoption increases, the use of 6to4 tunnels has declined, yet they remain a significant security concern in today’s Internet. This study investigates the real-world deployment of 6to4 tunnels, revealing their current scale, characteristics, and security implications. We identify open 6to4 relays in 216 countries and 13,114 autonomous systems, noting stable short-term counts but a long-term decline. We analyze the security of these nodes and find over 578k nodes vulnerable to address spoofing and packet injection. Additionally, we present several under-emphasized scenarios where open 6to4 nodes are abused, including leveraging services on 6to4 nodes as traffic amplifiers, circumventing restrictions using multiple 6to4 addresses, and connecting 6to4 nodes to render attacks untraceable.