Researcher, Zhejiang University
4 papers at NeurIPS 2025
We identify a smoothness-generalization dilemma in message passing that limits GNN universality across varying homophily and propose the Inceptive Graph Neural Network (IGNN), a universal framework to address the dilemma.
This paper explore the essence of message passing on heterophilic graphs from the perspective of model and design a method with the guidance of theoretical findings.
Learning the individualized treatment effects under the assumption of rank preservation.
