Chien-Yi Wang

Efficient Model Personalization in Federated Learning via Client-Specific Prompt Generation

2023 · arXiv (Cornell University)

Federated learning (FL) emerges as a decentralized learning framework which trains models from multiple distributed clients without sharing their data to preserve privacy. Recently, large-scale pre-trained models (e.g., Vision Transformer) have shown a strong capability …

MCPNet++: Interpretable Classification Models via Multi-Level Concept Prototypes

2026 · IEEE Transactions on Pattern Analysis and Machine Intelligence

Post-hoc and inherently interpretable methods have shown great success in uncovering the inner workings of black-box models, whether by examining them after training or by explicitly designing for interpretability. While these approaches effectively narrow the …