Junnan Liu

Rectifying LLM Thought from Lens of Optimization

G-reasoner: Foundation Models for Unified Reasoning over Graph-structured Knowledge

SituatedThinker: Grounding LLM Reasoning with Real-World through Situated Thinking

Reasoning as Meta-Learning: An Optimization Perspective to Decipher Long CoT Reasoning in LLMs

Coding Triangle: How Does Large Language Model Understand Code?

Learning with Interaction: Agentic Distillation for Large Language Model Reasoning

UnifiedVerifier: Unifying Paradigms in Automated LLM Evaluation

Steering Merged LLMs for Multilingual Reasoning with Coefficient Optimization

How Brittle is Agent Safety? Rethinking Agent Risk under Intent Concealment and Task Complexity

Rethinking Verification for LLM Code Generation: From Generation to Testing

Exploring the Limit of Outcome Reward for Learning Mathematical Reasoning

Combining Structure and Text: Learning Representations for Reasoning on Graphs