SEP 4, 2026 · PREPRINT
PACE: Propagation-Aware Collaborative Correction for One-Shot Personalized Federated Graph Learning
arXiv
This is a methodological preprint presenting an algorithm for federated graph learning with no peer review, clinical validation, or comparison to published baselines on standard benchmarks.
Reported
Communication efficiency (Rank-6)9.6–17.6% of dense tensor bytes
Datasets showing improvement5 out of 6
Rank parameter tested6