Shizhu He

DAComp: Benchmarking Data Agents across the Full Data Intelligence Lifecycle

Amplify Adjacent Token Differences: Enhancing Long Chain-of-Thought Reasoning with Shift-FFN

Beyond Hard and Soft: Hybrid Context Compression for Balancing Local and Global Information Retention

Towards Agentic Self-Learning LLMs in Search Environment

Dynamic Parametric Retrieval Augmented Generation

Probabilistic Uncertain Reward Model

The Zero-Step Thinking: An Empirical Study of Mode Selection as a Harder Early Exit Problem in Reasoning Models

From Instance Training to Instruction Learning: Task Adapters Generation from Instructions

Mastering Symbolic Operations: Augmenting Language Models with Compiled Neural Networks