<ul><li>Gossip Learning (GL) is a decentralized learning paradigm where users iteratively exchange and aggregate models with a small set of neighboring peers.</li><li>GRANITE is a framework for robust learning over sparse, dynamic graphs in the presence of Byzantine nodes.</li><li>GRANITE relies on a History-aware Byzantine-resilient Peer Sampling protocol (HaPS) to reduce adversarial influence over time.</li><li>Empirical results show that GRANITE maintains convergence with up to 30% Byzantine nodes and improves learning speed in sparser graphs.</li></ul>

GRANITE : a Byzantine-Resilient Dynamic Gossip Learning Framework

Discover more