Qingpei Guo

LLaVA-CMoE: Towards Continual Mixture of Experts for Large Vision-Language Models

FlattenGPT: Depth Compression for Transformer with Layer Flattening

HieraQuery: Bridging Multimodal Understanding and High-Quality Generation through Multi-Scale Query Learning

Rethinking the Impact of Heterogeneous Sublayers in Transformers