Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

PromptDyG: Test-Time Prompt Adaptation on Dynamic Graphs

About

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.

Guoguo Ai, Chaoxi Niu, Hui Yan, Joey Tianyi Zhou, Yew-Soon Ong, Guansong Pang• 2026

Related benchmarks

TaskDatasetResultRank
Temporal Link PredictionAS-733
MRR34.48
14
Temporal Link PredictionReddit-title
MRR0.4425
14
Temporal Link PredictionReddit-body
MRR39.06
14
Temporal Link PredictionUCI
MRR12.48
14
Temporal Link PredictionBitcoin-OTC
MRR20.04
14
Temporal Link PredictionBitcoin-Alpha
MRR17.32
14
Temporal Link PredictionUCI
AUROC0.9189
14
Temporal Link PredictionBitcoin-Alpha
AUROC94.65
14
Showing 8 of 8 rows

Other info

Follow for update