Xiaofeng Lin

I am Xiaofeng Lin, and I am currently pursuing my Ph.D. in Statistics and Data Science at the University of California, Los Angeles (UCLA). My work centers around the development of Large Language Models (LLMs) and generative models, with a strong emphasis on synthetic data generation for text and tables, as well as improving the alignment and safety of LLMs. I am dedicated to pushing the boundaries of AI to create solutions with significant practical applications.

My primary research interests include:

• Generative Models for Tabular Data: Designing and implementing foundation models and synthetic data generation techniques for tabular data. This involves developing methods to improve data privacy, address the challenges of limited data availability, and generate high-fidelity synthetic data for various applications.

• Large Language Models: Investigating methods for multi-turn optimization and chain-of-thought reasoning to enhance LLM performance and interpretability. I also explore alignment and safety in LLM and multi-LLM agent collaboration, working towards building trustworthy and ethical AI systems.