Associate Professor, Fudan University
5 papers at NeurIPS 2025
We propose a bilevel framework that jointly learns layer-wise sparsity and low-rank structure via truncated Gaussian sampling and efficient matrix approximation, achieving better performance for compressing LLMs.
We employ contrastive learning to extract complete point cloud structures from partial (incomplete) point clouds for guiding point cloud completion, achieving state-of-the-art (SOTA) results in the field of self-supervised point cloud completion.
