Shibo Hao

LLM Pretraining with Continuous Concepts

Emergence of Superposition: Unveiling the Training Dynamics of Chain of Continuous Thought

DenseMixer: Improving MoE Post-Training with Precise Router Gradient

Concise Reasoning in the Lens of Lagrangian Optimization

Learning Modal-mixed Chain-of-thought Reasoning with Latent Embedding

Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thought

Training Large Language Models to Reason in a Continuous Latent Space

Flow of Reasoning: Training LLMs for Divergent Reasoning with Minimal Examples

Training Large Language Model to Reason in a Continuous Latent Space

Flow of Reasoning: Training LLMs for Divergent Problem Solving with Minimal Examples