Abstract
Deep graph learning models deployed in real-world systems often need to cope with non-stationary environments, where the underlying graph distribution drifts continually over time. Prevailing solutions rely on training auxiliary generative modules to synthesize memory graphs for cross-domain adaptation, which incurs substantial computational overhead and scales poorly under prolonged distribution shifts. We argue that a more economical path exists: rather than generating memory, one can crystallize it. To this end, we propose Efficient Memory Crystallization (EMC), a training-free test-time framework that distills each incoming graph domain into a compact, semantically faithful memory through a closed-form solution to a memory-oriented distribution-matching objective, thereby eliminating redundant domain information under continual covariate shifts. To preserve both generalizability and adaptability as the model traverses a long sequence of target domains, EMC further models inter-domain dependencies through state-evolving memories and admits a theoretically grounded, tighter generalization error bound than direct adaptation. Extensive experiments demonstrate the superior performance of EMC over state-of-the-art baselines on graphs under non-stationary distribution shifts, while reducing average runtime by 87.4% and GPU memory consumption by 92.4% relative to the recent competitor, making continual graph adaptation practical at scale.
Keywords
Subject
Publication details
- Journal
- Not available
- Open access
- Green open access
Cite this article
APA 7
Hou, Y., Liu, R., Su, Y., Wu, J., & Xu, K. (2026). Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts. https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts
MLA 9
Hou, Yue, et al. "Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts." https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts.
Chicago (author–date)
Hou, Yue, Ruomei Liu, Yingke Su, Junran Wu, and Ke Xu. 2026. "Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts." https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts.
Harvard
Hou, Y., Liu, R., Su, Y., Wu, J. and Xu, K. (2026) 'Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts', Available at: https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts.
Vancouver
Hou Y, Liu R, Su Y, Wu J, Xu K. Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts. https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts
IEEE
Y. Hou, R. Liu, Y. Su, J. Wu, and K. Xu, "Efficient Memory Crystallization for Graph Learning under Non-Stationary Distribution Shifts," https://omanscience.com/en/articles/efficient-memory-crystallization-for-graph-learning-under-non-stationary-distribution-shifts.