Hi! I am a third-year Ph.D. student at UNC Chapel Hill, advised by Prof. Mohit Bansal. My research targets enabling effective exploration for improved LLM reasoning. By structuring interaction (across agents) and adapting human cognitive behavior (within agents), my work aims to improve exploration, self-correction, and generalization, making reasoning more efficient, adaptive, and less dependent on heavy supervision. I received my master’s degree from National Taiwan University, advised by Dr. Lun-Wei Ku and Prof. Hsin-Hsi Chen. When I was an undergraduate, I had the pleasure of working closely with Prof. Cheng-Te Li.

Recent News

• May 2026: Skill-MoE is accepted to ICML 2026.
• Apr 2026: I am joining Microsoft Research as a research intern this summer!
• Mar 2026: Check out Cog-DRIFT, our new preprint on arxiv!
• Jan 2026: NuRL is accepted to ICLR 2026.
• Sep 2025: Check out NuRL, our new preprint on arxiv!
• Aug 2025: MAgICoRE is accepted to EMNLP 2025.
• May 2025: I am joining Salesforce AI Research as a research intern this summer!
• Apr 2025: Presented Reverse Thinking Makes LLMs Stronger Reasoners at MPI+UCLA.
• Mar 2025: Presented Reverse Thinking Makes LLMs Stronger Reasoners at Google DeepMind.
• Mar 2025: Check out Skill-MoE, our new preprint on arxiv!
• Jan 2025: RevThink is accepted to NAACL 2025.
• Nov 2024: Check out RevThink, our new preprint on arxiv!
• Sep 2024: Check out MAgICoRE, our new preprint on arxiv!
• Jul 2024: Check out System-1.x, our new preprint on arxiv!
• Jun 2024: I am joining Google Cloud AI as a Student Researcher this summer!
• May 2024: ReConcile is accepted to ACL 2024.
• May 2024: MAGDi is accepted to ICML 2024.
• Feb 2024: Check out MAGDi, our new preprint on arxiv!
• Oct 2023: We have a paper accepted to EMNLP 2023.
• Sep 2023: Check out ReConcile, our new preprint on arxiv!
• May 2023: We have two papers accepted to ACL 2023.
• Apr 2023: I will be joining MURGe-Lab and begin my Ph.D. journey!