Chun-Liang Li

Chun-Liang Li is a research scientist at Apple MLR and an affiliate assistant professor at the Paul G. Allen School of Computer Science & Engineering, University of Washington. His work bridges machine learning theory with practical applications in computer vision and natural language processing, focusing on efficient model training and representation learning.

His educational background includes:

• Ph.D. in Machine Learning from Carnegie Mellon University (2014-2019), supervised by Prof. Barnabás Póczos
• B.S. and M.S. in Computer Science and Information Engineering from National Taiwan University (2008-2013), supervised by Prof. Hsuan-Tien Lin

Li's research centers on generative models and representation learning, with significant contributions to document understanding (FormNet series), multimodal systems (Pic2word), and large language model efficiency. His work consistently addresses real-world challenges like reducing training costs while maintaining performance, as seen in distillation techniques and synthetic data optimization.

Analysis of his 2022-2024 publications reveals three dominant trends: (1) LLM efficiency through curriculum training and model updating, (2) structural document understanding via graph-based methods, and (3) multimodal representation learning for vision-language tasks. These reflect his cross-cutting approach to improving model scalability and applicability.

His scientific recognition includes:

• IBM Ph.D. Fellowship (2018)
• Best student paper runner-up at IJCAI (2017)
• Double first-place wins in KDD Cup Tracks (2011, 2013)

While specific grant details aren't listed, his award-winning KDD Cup performances and extensive publication record suggest strong funding support. He collaborates widely with students and researchers, though formal advisees aren't specified. His current roles at Apple MLR and UW position him at the industry-academia interface for cutting-edge AI development.

At Apple, Li contributes to the Machine Learning Research group's core vision-language projects, while his UW affiliation enables academic mentorship and cross-institutional collaboration on foundational ML research.