Li Sun

Associate Professor@Beijing University of Post and Telecommunications·中国·OpenReview

研究方向

AI for Science · Molecule Design · Stochastic Differential Equation · Ordinary Differential Equation · Physics-informed Neural Network · Neural Operator · Deep Clustering · Unsupervised Learning · Graph Neural Network · Graph Machine Learning · Riemannian Geometry · Geometric Deep Learning · Social Network Analysis · Computational Social Science

Li Sun

Multi-Domain Riemannian Graph Gluing for Building Graph Foundation Models

Target Before You Perturb: Enhancing Locally Private Graph Learning via Task-Oriented Perturbation

Resource Consumption Red-Teaming for Large Vision-Language Models

Flow-based Automatic Neural Operator with Hard Physical Constraints

FitLight: Federated Imitation Learning for Plug-and-Play Autonomous Traffic Signal Control

ChiFlow: Torsional Asymmetry Flow Matching For Chirality-Aware Protein Backbone Generation

Diagnosing Manifold Collapse: Intervention-Based Topology in Low-Rank RNNs

Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Models