Abstract: Sample alignment is recognized as a vital component of vertical federated learning, which facilitates the integration of differential samples and high-quality model training. In this trend, ...