Hengyu Fu
3
papers
6
total citations
papers (3)
Diffusion Transformer Captures Spatial-Temporal Dependencies: A Theory for Gaussian Process Data
ICLR 2025arXiv
3
citations
Learning Hierarchical Polynomials of Multiple Nonlinear Features
ICLR 2025arXiv
3
citations
What can a Single Attention Layer Learn? A Study Through the Random Features Lens
NeurIPS 2023arXiv
0
citations