Haonan Li works on large language models with a focus on post-training, data engineering, and reasoning capabilities. His research bridges open-source development and production-ready foundation models.