Carnegie Mellon University

Chenyan Xiong

Chenyan Xiong

Associate Professor, Language Technologies Institute

Address
5000 Forbes Avenue
Pittsburgh, PA 15213

Chenyan Xiong is an Associate Professor at Carnegie Mellon University's Language Technologies Institute (LTI), where his research focuses on foundation and large language models (LLMs), information retrieval, and healthcare AI. Dr. Xiong has developed innovative methods in Neural IR, dense retrieval, and efficient foundation model pretraining with both academia and industry influences, as well as datasets like ClueWeb22 and Conversational Search. His research has been widely recognized in premier conferences such as ICLR, ACL, and SIGIR.

Dr. Xiong earned his Ph.D. and M.S. in Language Technologies from Carnegie Mellon University, after completing his M.Eng. at the Chinese Academy of Sciences and his B.Eng. at Wuhan University. Before returning to CMU as a faculty member, he held research leadership roles at Microsoft Research Redmond, where he pioneered solutions for information retrieval and large-scale language model pretraining. He is the recipient of several prestigious honors, including the inaugural SIGIR Early Career Award.

At CMU, Dr. Xiong teaches and develops courses such as "Large Language Models and Applications," fostering a hands-on understanding of state-of-the-art LLMs. He actively mentors students across multiple programs and serves as a leader in the research community through roles like tutorial chair at SIGIR and associate editor for TPAMI. His dedication to advancing the capabilities and accessibility of LLMs continues to influence both academia and industry.

  • Data-Centric Foundation and Language Models: Build efficient and capable foundation and language models through data-centric approaches.
  • Embedded Learning: Represent the rich information in documents, images, videos, and various modalities into an embedding vector to empower various information retrieval scenarios.
  • New GenAI-Enabled Scenarios: Explore new application scenarios enabled by Generative AI technologies