Xianzhi Du

Adaptive Thinking: Large Language Models Know When to Think in Latent Space

MANZANO: A Simple and Scalable Unified Multimodal Model with a Hybrid Vision Tokenizer

On the Efficiency of Structured Pruning in Small Language Model Pretraining

MM1.5: Methods, Analysis & Insights from Multimodal LLM Fine-tuning

Fantastic Experts and How to Find Them: A Multi-Dimensional Study for Experts-Level Sparsification in Mixture-of-Experts