Prof. Xuan Wang
Virginia Tech
Towards Effective and Efficient Language Model Agents for Structured Knowledge Extraction
Large language models have demonstrated remarkable capabilities across both scientific discovery and societal applications. Yet, their immense size and proprietary nature often limit accessibility, reproducibility, and broad adoption. In this talk, I will present our efforts to develop small, open-source, multi-modal language model agents that can reason, plan, and act across diverse contexts. I will outline strategies for building compact but powerful models, integrating multi-modal data, and enabling collaboration through multi-agent systems to tackle complex tasks. A particular focus will be on leveraging multi-agent interactions for structured knowledge extraction from massive text, which provides a natural bridge to graph and relational representations central to many scientific domains.
Biography
Dr. Xuan Wang is an Assistant Professor in the Department of Computer Science at Virginia Tech. Her research interests are in natural language processing, data mining, AI for sciences, and AI for healthcare. Her current research topics include natural language understanding with limited supervision, complex reasoning and planning with language model agents, and multi-modal science foundation models. She was a recipient of the NSF CAREER Award 2025, Nvidia Academic Grant 2025, Cisco Research Award 2025, NSF NAIRR Pilot Award 2024 - 2025, and NAACL Best Demo Paper Award 2021. She received a Ph.D. degree in Computer Science, an M.S. degree in Statistics, and an M.S. degree in Biochemistry from the University of Illinois Urbana-Champaign in 2022, 2017, and 2015, respectively, and a B.S. degree in Biological Science from Tsinghua University in 2013.
More information can be found at https://xuanwang91.github.io/