PromptDyG: Test-Time Prompt Adaptation on Dynamic Graphs
2026-06-22 • Machine Learning
Machine Learning
AI summaryⓘ
The authors study how to better understand changing systems represented as series of graphs over time. They point out that existing methods train once and then don't adapt to new information, which can miss important changes. To fix this, they propose PromptDyG, which continuously updates itself using patterns learned without needing labels, making it more flexible as the system evolves. Their method is supported by theory and tested on multiple datasets, showing improved predictions compared to previous approaches.
dynamic graphsdiscrete-time dynamic graphsonline learningprompt adaptationunsupervised learningentropy minimizationgraph snapshotsdistribution shiftrepresentation learning
Authors
Guoguo Ai, Chaoxi Niu, Hui Yan, Joey Tianyi Zhou, Yew-Soon Ong, Guansong Pang
Abstract
Activities in numerous evolving systems can be represented as dynamic graphs in snapshot form at different time intervals, i.e., discrete-time dynamic graphs (DTDGs). Existing methods show impressive advances in capturing historical temporal evolution patterns in DTDGs, but they focus on addressing an offline learning setting, where models are trained using historical snapshots once and then evaluated to all subsequent graph snapshots without further updating. This fails to capture 1) the nature of evolving complexities across graph snapshots and 2) the distribution shift in the testing graph snapshots. To address these problems, we propose PromptDyG, a novel framework that leverages unsupervised test-time Prompt adaptation for Dynamic Graph learning under a live-update online setting. The key insight is that an expressive dynamic graph prompt can be learned on a frozen backbone via minimization of feature-wise, label-free entropy to efficiently and continuously model the evolving patterns. We show theoretically that this unsupervised prompt adaptation can guarantee a larger similarity margin between positive and negative pairs, facilitating more accurate dynamic predictions. It is further confirmed by our extensive empirical results on six benchmark datasets that show consistent and significant improvements of PromptDyG over state-of-the-art baselines.