Associate Professor, University of Connecticut
3 papers at NeurIPS 2025
A graph topology-oriented prompting framework
We treat substructures as graph tokens to enable the scale transformer pre-training on graphs.