TY  - RPRT
TI  - Convex and Non-convex Federated Learning with Stale Stochastic Gradients: Diminishing Step Size is All You Need
AU  - Xinran Zheng
AU  - Tara Javidi
AU  - Behrouz Touri
PY  - 2026
UR  - https://arxiv.org/abs/2603.02639
ID  - 2603.02639
ER  -