Zhiqin Yang

Data Selection for LLM Reinforcement Learning with Improved Gradient Alignment

SurstSplat: Dynamic Surgical Gaussian Reconstruction with Spatiotemporal Graph Matching

MimicDreamer: Aligning Human and Robot Demonstrations for Scalable VLA Training

FedGPS: Statistical Rectification Against Data Heterogeneity in Federated Learning

GaussianClin: Multimodal Featured Gaussian Splatting for Dynamic Clinical Videos