Recommender systems are essential for filtering content to match user preferences. However, traditional recommender systems often suffer from biases inherent in the data, such as popularity bias. These biases, particularly those stemming from latent confounders, can result in inaccurate recommendations and reduce both the diversity and effectiveness of the system. Existing debiasing methods for recommender systems, however, either fail to account for latent confounders or rely on predefined instrumental variables (IVs). To address this research gap, we propose a novel causality-based recommendation algorithm, Data-driven IV representation learning for debiasing in Recommender System (DIVRS), which enables the learning of IV representation directly from user-item interaction data. By leveraging the learned IV representation, DIVRS decomposes user behaviour into causal and confounding relationships to address potential bias in recommender systems. Additionally, we introduce Orthogonal Promotion Regularisation (OPR) for DIVRS to address the problem that Graph Convolutional Networks (GCNs) amplify bias. We also propose a variant of GCNs for DIVRS, called DIVRS-GCN. Experimental results on the Douban-Movie and Movielens-10M datasets demonstrate that both DIVRS and DIVRS-GCN effectively mitigate confounding bias while outperform the state-of-the-art methods in recommendation performance. For example, on both datasets, our DIVRS and DIVRS-GCN improve Recall@20 by up to 10.98 %. This validates their effectiveness and robustness. Our approaches improve recommendation accuracy while delivering more balanced and diverse suggestions, effectively addressing the limitations of existing IV-based recommender systems.

Visual analysis of spatiotemporal density distributions is crucial for understanding spatiotemporal dynamics. However, existing methods suffer from visual occlusion and information loss when simultaneously displaying multiple density distributions. We present Strunkmap as an abstract approach to address these challenges. We introduce anisotropic kernel density estimation to enhance the accuracy of density generation. We extract the trunks of density distributions to identify the overall spatial patterns. Path scanning and trunk-outline matching strategies are employed to preserve local spatial structure. We design a stacked trunk plot that enables lossless density representation while conserving substantial screen space. Based on the visual design, Strunkmap integrates multiple heatmaps within a single map to effectively display temporal evolution of density distributions without visual occlusion. Ablation studies and comparative experiments validate the superiority of Strunkmap in accuracy and efficiency for hotspot identification and trend exploration. Theoretical analysis demonstrates Strunkmap's scalability, which we further verify through large-scale spatiotemporal data visualization. Color encoding schemes and scaling ratios are discussed to illustrate the flexibility. Our evaluations with user feedback demonstrate that Strunkmap is a viable solution with significant potential to real-world applications.

With the increasing security issues in blockchain, smart contract vulnerability detection has become a research focus. Existing vulnerability detection methods have their limitations: 1) Static analysis methods struggle with complex scenarios. 2) Methods based on specialized pre-trained models perform well on specific datasets but have limited generalization capabilities. In contrast, general-purpose Large Language Models (LLMs) demonstrate impressive ability in adapting to new vulnerability patterns. However, they often underperform on specific vulnerability types compared to methods based on specialized pre-trained models. We also observe that explanations generated by generalpurpose LLMs can provide fine-grained code understanding information, contributing to improved detection performance. Inspired by these observations, we propose SAEL, a LLMbased framework for smart contract vulnerability detection. First, we design prompts targeting specific smart contract vulnerabilities to guide general-purpose LLMs in detecting vulnerabilities and providing explanations. The detection results generated by LLMs serve as prediction features. Then, we employ prompt-tuning on CodeT5 and T5 respectively to process contract code and explanations, enhancing model performance on specific tasks. To leverage the strengths of each component, we introduce Adaptive Mixture-of-Experts, a dynamic architecture for smart contract vulnerability detection. This mechanism dynamically adjusts feature weights through a Gating Network, which selects the most relevant features by applying TopK filtering and Softmax normalization, and a Multi-Head Self-Attention mechanism, which enhances cross-feature relationships by processing multiple attention heads in parallel. This design ensures that prediction results for LLMs, explanation features, and contract code features are effectively integrated through gradient optimization. The loss function focuses on the independent prediction performance of each feature and the overall performance of weighted predictions. Experimental results show that SAEL outperforms existing methods in detecting various vulnerabilities.

It is often challenging to identify a valid instrumental variable (IV), although the IV methods have been regarded as effective tools of addressing the confounding bias introduced by latent variables. To deal with this issue, an Interaction-Data-guided Conditional IV (IDCIV) debiasing method is proposed for Recommender Systems, called IDCIV-RS. The IDCIV-RS automatically generates the representations of valid CIVs and their corresponding conditioning sets directly from interaction data, significantly reducing the complexity of IV selection while effectively mitigating the confounding bias caused by latent variables in recommender systems. Specifically, the IDCIV-RS leverages a variational autoencoder (VAE) to learn both the CIV representations and their conditioning sets from interaction data, followed by the application of least squares to derive causal representations for click prediction. Extensive experiments on two real-world datasets, Movielens-10M and Douban-Movie, demonstrate that IDCIV-RS successfully learns the representations of valid CIVs, effectively reduces bias, and consequently improves recommendation accuracy.

The cell cycle is a fundamental process in systems biology, and Ordinary Differential Equations (ODEs) constitute a valuable tool for describing cell cycle dynamics. A major challenge in this field is estimating unknown parameters in these equations from sparse biological data. Recently, a class of machine learning methods, namely Physics-Informed neural networks (PINNs), has been proposed for solving ODEs and the corresponding parameter estimation problem. PINNs incorporate physical laws into the loss function of a neural network, with these losses evaluated at a set of scattered spatio-temporal points (called residual points). In this work, we propose a new point-wise weighting method for improving the accuracy of loss evaluation at these residual points, termed Time Scale Gradient Aggregation Self-Attention (TGSA). TGSA assigns dynamic weights to residual points by leveraging time-scale relationships and gradient information between state variables, capturing the internal dependencies among residuals. Our experimental results demonstrate the superior performance of TGSA in parameter estimation for four distinct cell cycle models.

The automation of code review activities has emerged as a critical research focus for optimizing development efficiency while ensuring code quality. While recent advancements in Large Language Models (LLMs) have shown promise, existing approaches predominantly isolate the three core code review tasks—review necessity prediction, review comment generation, and code refinement, overlooking their valuable interdependencies. Empirical analysis reveals that isolated-trained comment-generation models often produce superficial comments (e.g., “Undefined ‘userInput’”) due to insufficient understanding of defect patterns, which is what necessity prediction tasks precisely target. Recent efforts to model interdependencies through knowledge distillation remain constrained by static framework designs.To address these challenges, we present MoE-Reviewer, which adopts the Mixture-of-Experts (MoE) framework on the LLaMA model to tackle the interdependence of code review tasks. MoE-Reviewer enables collaborative modeling for the three tasks mentioned above. By integrating dynamic coordination routing strategies and fine-grained expert mechanisms, MoE-Reviewer facilitates effective knowledge sharing across tasks while mitigating parameter interference. Evaluations conducted on the CodeReviewer dataset demonstrated that MoE-Reviewer outperforms existing methods, achieving state-of-the-art performance with an F1-score of 73.2% and improving the BLEU score for review comment generation by 5.32 to 11.62. Additionally, routing analysis further validates the effectiveness of our approach.

The task of issue resolving aims to modify a codebase to generate a patch that addresses a given issue. However, most existing benchmarks focus almost exclusively on Python, making them insufficient for evaluating Large Language Models (LLMs) across different programming languages. To bridge this gap, we introduce a multilingual issue-resolving benchmark, called Multi-SWE-bench, covering 8 languages of Python, Java, TypeScript, JavaScript, Go, Rust, C, and C++. In particular, this benchmark includes a total of 2,132 high-quality instances, carefully curated by 68 expert annotators, ensuring a reliable and accurate evaluation of LLMs on the issue-resolving task. Based on human-annotated results, the issues are further classified into three difficulty levels. We evaluate a series of state-of-the-art models on Multi-SWE-bench, utilizing both procedural and agent-based frameworks for issue resolving. Our experiments reveal three key findings: (1) Limited generalization across languages: While existing LLMs perform well on Python issues, their ability to generalize across other languages remains limited; (2) Performance aligned with human-annotated difficulty: LLM-based agents' performance closely aligns with human-assigned difficulty, with resolution rates decreasing as issue complexity rises; and (3) Performance drop on cross-file issues: The performance of current methods significantly deteriorates when handling cross-file issues. These findings highlight the limitations of current LLMs and underscore the need for more robust models capable of handling a broader range of programming languages and complex issue scenarios.

Transformers have been widely regarded as a promising direction for breaking through the performance bottlenecks of Graph Neural Networks (GNNs), primarily due to their global receptive fields. However, a recent empirical study suggests that tuned classical GNNs can match or even outperform state-of-the-art Graph Transformers (GTs) on standard node classification benchmarks. Motivated by this fact, we deconstruct several representative GTs to examine how global attention components influence node representations. We find that the global attention module does not provide significant performance gains and may even exacerbate test error oscillations. Consequently, we consider that the Transformer is barely able to learn connectivity patterns that meaningfully complement the original graph topology. Interestingly, we further observe that mitigating such oscillations enables the Transformer to improve generalization in GNNs. In a nutshell, we reinterpret the Transformer through the lens of graph spectrum and reformulate it as a global-aware graph filter with band-pass characteristics and linear complexity. This unique perspective introduces multi-channel filtering constraints that effectively suppress test error oscillations. Extensive experiments (17 homophilous, heterophilous graphs) provide comprehensive empirical evidence for our perspective. This work clarifies the role of Transformers in GNNs and suggests that advancing modern GNN research may still require a return to the graph itself.

Graph Anomaly Detection (GAD) has gained significant attention in areas such as financial risk control and social network security, becoming a critical research problem. Vanilla Graph Neural Networks (GNNs), a popular method for graph modeling, are known to perform poorly in GAD due to the assumption of homophily preferences. This article argues that the issue lies in the insufficient feature extraction ability caused by their single filtering property (low-pass filtering) and revealing the effectiveness of multi-band filtering to deal with GAD. From this, we note two other overlooked issues: (1) How can multi-band band-pass filtering further fuse multi-scale neighborhood information? (2) Adaptation between raw attributes of nodes and graph filters (graph topology). The former bridges the respective advantages of spectral domain and spatial domain, and the latter is an important bottleneck for the encoding capacity of the filters. To address these, we propose a new GAD method, Graph Perturbed Networks (GraphPN). Each hidden layer of GraphPN is a band-pass filter, enabling multi-band and multi-scale filtering through simple stacking and skip connections. We analyze its spectral locality and spatial locality to provide theoretical support. Additionally, GraphPN is supplemented with a tailored feature activation module to complete the adaptation of the above two. This module readjusts node indices and decouples graph convolution, introducing rich topological information to node attributes. In addition to further enhancing detection performance, another possibly counter-intuitive effect is that the distinguishability of the two classes of nodes is improved even before filtering. The proposed method performs well in real-world datasets compared with the current state-of-the-art baselines, which fully demonstrates its superiority. Codes are available at https://github.com/Thankstaro/GraphPN .

The efficient operation of the transportation system plays a vital role in personal mobility, economic development, environmental sustainability, and stability of social operations. Traffic flow prediction is a critical task, and accurate traffic prediction can reduce traffic congestion and improve the operational efficiency of the transportation system. Existing models use given spatial adjacency graphs and sophisticated mechanisms to model spatial and temporal correlations and obtain good results. However, most of them rely on predefined graph structures or only consider the dynamic changes of temporal relations and ignore the dynamic changes of spatial relations. In order to solve these problems, this study proposes a spatio-temporal dynamically fused graph-convolutional network(STDFGCN) for traffic prediction. Unlike traditional models, our model does not rely on predefined graph structure but pays more attention to the dynamic changes of spatial relations and has a more vital adaptive ability to effectively capture and fuse the dynamically changing features of temporal relations and spatial relations. The model can understand the complexity of the urban transportation system more comprehensively, improving the prediction performance and enhancing traffic prediction ability. Extensive experiments have been conducted on three real motorway traffic datasets, and the results show that this paper is better than the current state-of-the-art related methods.

Decentralized Finance (DeFi) ecosystem has grown rapidly in the past few years. In the DeFi ecosystem, flash loan is a novel type of uncollateralized loan with nearly negligible lending costs. Malicious attackers can easily borrow a large number of crypto assets, and utilize them to disrupt the price of crypto assets to make a profit. Many flash loan based price manipulation attacks have been reported recently, and caused immense economic losses, e.g., 30 million USD in a single attack. In this paper, we conduct an empirical study on real-world flash loan based attacks in the past two years and present three attack patterns for price manipulation attacks. Then, we propose an approach, LeiShen, to automatically detect price manipulation attacks with asset transfers. We evaluate LeiShen on the first 14,500,000 blocks in Ethereum, and detect 180 attacks with a precision of 78.9%. Among our newly-found attacks, the severest attack has caused a total loss of more than 6.1 million USD.