TH-GNN is a heterogeneous temporal graph neural network designed to detect shilling attacks generated by large language model (LLM) agents. It combines a two‑layer Heterogeneous Graph Transformer with per‑type and per‑relation attention, learnable sinusoidal temporal encodings, cross‑modal attention that fuses user embeddings with frozen RoBERTa representations of reviews and item descriptions, and a GRU that models log inter‑arrival times. Across five attack families and four benchmark datasets, TH‑GNN achieves a grand‑mean F1 score of 0.870, surpassing the best text‑only baseline on Agent4SR attacks by 10.9 percentage points and 11.5 percentage points at the lowest injection rate.
arXiv:2608. 12441v1 Announce Type: cross Abstract: Deep learning detectors for anomalies in dynamic graphs have reached strong accuracy, yet they remain opaque: when an edge is flagged, the analyst receives a score but no reason.
The paper reinterprets graph neural networks (GNNs) as retrieval-augmented models, where each layer uses an MLP on a node representation and a permutation‑invariant summary of retrieved graph context instead of traditional message passing. It introduces RTA, a lightweight MLP‑based framework that replaces structural message passing with label‑aware retrieval and propagation, and provides theoretical links to softmax‑attention message passing and robustness to mis‑retrieved outliers. Experiments on text‑attributed graph benchmarks demonstrate that RTA matches or surpasses strong GNN and graph LLM baselines while improving efficiency and robustness.
The paper evaluates privacy risks in graph neural networks enhanced by large language models (LLMs). Using a five‑stage framework, the authors test six real‑world text‑attributed graph datasets with 42 model configurations and six privacy attack methods across link, label, and membership inference threats. Results show that LLM‑enhanced GNNs are more vulnerable than shallow baselines, with semantic enrichment amplifying exploitable signals, and that differential privacy can reduce risk but at a significant cost to utility.
FoundAna is a GNN‑assisted foundation model designed for graph anomaly detection across diverse datasets. It combines a GNN component with a transformer encoder enhanced by four positional encodings to capture both local and global structure, using reconstruction errors as anomaly scores. Experiments on nine benchmark datasets from financial, social, and citation networks show that FoundAna consistently outperforms state‑of‑the‑art baselines.
arXiv:2606. 17109v1 Announce Type: cross Abstract: Given their effectiveness in modeling the relational structure among network traffic flows, graph neural networks (GNNs) have been widely adopted in network intrusion detection systems (NIDSs).
arXiv:2607. 03350v1 Announce Type: cross Abstract: Malicious Python packages have become a major threat to software supply chain ecosystems due to the widespread adoption of open-source repositories such as PyPI.
arXiv:2606. 28134v1 Announce Type: cross Abstract: Graph-based fraud detection is essential for safeguarding large-scale transaction systems, where undetected anomalies may lead to substantial financial losses and security risks.
TTGBench is a new benchmark for temporal graph learning that evaluates both structural evolution and semantic drift in text‑attributed graphs. It includes six real‑world, text‑rich datasets with dual volatility and supports multi‑class and multi‑label temporal node classification, addressing gaps left by existing benchmarks. A comprehensive evaluation of 17 state‑of‑the‑art methods shows a clear divide: TGNNs excel at structural prediction but struggle with semantic tracking, while LLM‑based models perform better on semantic tasks but lag in structural prediction.
arXiv:2607. 19674v1 Announce Type: cross Abstract: Federated Graph Neural Networks (FedGNNs) are highly vulnerable to backdoor poisoning, yet existing defenses typically rely on rule-based approaches that lack semantic understanding, making them vulnerable to stealthy triggers and harmful to benign structures.
The paper introduces MOSAIC, a large adversarial benchmark for detecting AI-generated text, and presents NeuroStat, a new framework that combines token‑level probabilistic logits with deep semantic hidden states from a single language model. NeuroStat fuses these signals via Macro‑State Residual Modulation and uses orthogonal and contrastive losses to learn complementary representations. Experiments show that NeuroStat outperforms existing methods on MOSAIC, achieving superior robustness against adversarial attacks.
arXiv:2606. 08067v1 Announce Type: new Abstract: Graph neural networks (GNNs) are widely deployed on relational data, yet they can leak sensitive or proprietary information about the training graph adjacency, e.
Kernel-Complexity Edge Sanitization (KCES) is a training‑free, model‑agnostic defense for Graph Neural Networks that identifies and removes edges with high Kernel‑Complexity (KC) scores, which are indicative of structural influence on the graph kernel complexity metric. KCES leverages a theoretical upper bound on GNN test error derived from the graph Gram matrix to compute edge‑specific KC scores, pruning edges that are empirically enriched with adversarial perturbations. The method is computationally efficient, scalable to large graphs, and consistently outperforms representative robust baselines across diverse attack settings without requiring retraining.