2025/10/10 by Abigail J. Hayes, Tobias Schumacher, Hayes, Abigail J. +3 · 1 voice · 1 citation
Computer Science · Psychology · #Advanced Graph Neural Networks #Benchmark (surveying) #Data modeling #Deep learning #Feature learning #Graph #Machine Learning in Healthcare #Mental Health Research Topics #Representation (politics) #Simple (philosophy)
paper · pdf · doi:10.48550/arxiv.2510.09416
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2025/10/10 · openalex created_date 2025/10/14 · openalex updated_date 2026/08/01
Learning on temporal graphs has become a central topic in graph representation learning, with numerous benchmarks indicating the strong performance of state-of-the-art models. However, recent work has raised concerns about the reliability of benchmark results, noting issues with commonly used evaluation protocols and the surprising competitiveness of simple heuristics. This contrast raises the question of which characteristics of the underlying graphs temporal graph learning models actually use to form their predictions. We address this by systematically evaluating eight models on their ability to capture eight fundamental characteristics related to the link structure of temporal graphs. These include structural characteristics such as density, temporal patterns such as recency, and edge formation mechanisms such as homophily. Using both synthetic and real-world datasets, we analyze how well models learn these characteristics. Our findings reveal a mixed picture: models capture some characteristics well but fail to reproduce others. With this, we expose important limitations. Overall, we believe that our results provide practical insights for the application of temporal graph learning models and motivate more interpretability-driven evaluations in graph learning research.