Li Yi

Latest

• VILA-U: a Unified Foundation Model Integrating Visual Understanding and Generation
• SparseViT: Revisiting Activation Sparsity for Efficient High-Resolution Vision Transformer