Hengyi Cai

CurES: From Gradient Analysis to Efficient Curriculum Learning for Reasoning LLMs

Solving the Granularity Mismatch: Hierarchical Preference Learning for Long-Horizon LLM Agents

Staying in the Sweet Spot: Responsive Reasoning Evolution via Capability-Adaptive Hint Scaffolding

Long-Attention Weaving: Synthesizing Long-Context Data with High-Quality Short Data

From Exploration to Mastery: Enabling LLMs to Master Tools via Self-Driven Interactions