Xiaowen Jiang presents oral paper at ICLR 2026

Xiaowen Jiang had the pleasure of presenting our oral paper Non-Convex Federated Optimization under Cost-Aware Client Selection at ICLR 2026 in Rio de Janeiro, joint work with Anton Rodomanov and Sebastian U. Stich.

The work studies how federated learning systems can make better use of limited communication and heterogeneous client resources, helping make distributed machine learning more efficient and reliable in practical settings.

We were also happy to see good representation from our group and collaborators throughout the conference, with a total of six poster presentations and many inspiring discussions.