Associate Professor, CMU, Carnegie Mellon University
1 paper at NeurIPS 2025
We propose a parameter-efficient fine-tuning method that improves the performance of LoRA in the presence of computational and data heterogeneity in FL.