cv

PhD

Texas A&M University, College Station, Texas

Computer Science

Research Assistant

Stanford University

Department of Computer Science, Zou's Group, Choi's lab. Advisor: Prof. James Zou and Prof. Yejin Choi

• Agentic Scientific LLM Post-training

Research Assistant

University of Waterloo

Department of Computer Science, TIGER-AI-Lab. Advisor: Prof. Wenhu Chen

• Agentic Tool-Use LLMs through RL
• Propose a novel agentic async tool-use RL training framework
• Achieve strong performance across diverse benchmarks, including math and search tasks
• Open-source tool-agent training framework Verl-Tool (500+ stars now) and submit work to ICLR 2026

Machine Learning Researcher

Kuaishou

Haidian District, Beijing

• Generative Personalized Re-ranking Recommendation
• Develop an end-to-end generative training framework for re-ranking recommendations powered by LLM, enhancing Recommendation System generalization and personalization
• Deliver significant online gains on Kuaishou (300 M+ DAUs) and recognized as an excellent LR (launch review)
• Accepted by CIKM 2025

Research Assistant

Emory University

Department of Computer Science. Advisor: Prof. Liang Zhao

• LLMs for Textual Graph Mining
• Propose a novel framework for link prediction on textual-edge graphs by jointly leveraging graph topology and semantic information. The method integrates coherent document composition and LLM-enhanced self-supervised training to equip GNNs with language understanding
• Conduct extensive experiments on four real-world datasets, demonstrating that our method boosts the performance of general GNNs and achieves competitive results compared to edge-aware GNNs
• Accepted by NeurIPS 2024