Efficient Model Personalization in Federated Learning via Client-Specific Prompt Generation | IEEE Conference Publication | IEEE Xplore