Hi everyone, I am teaching the NLP course in CS@VT this semester. We welcome you to join us for a series of invited talks from external speakers. The first speaker is Dr. Yu Zhang from CS at Texas A&M. Time: Tuesday, 4/14/26, 11 am - 12:15 pm Location: https://virginiatech.zoom.us/j/8682759193 Below please find the talk information: Title: Assisting Scientific Research with Structure-Aware Large Language Models Abstract: Large language models (LLMs) are powerful yet prone to hallucinations and are limited in handling complex reasoning for knowledge-intensive tasks. In this talk, I will present our work on augmenting LLMs with structured knowledge to support scientific research at multiple stages. First, I will discuss methods for answering researchers’ queries, including chain-of-thought reasoning and reinforcement learning over scientific knowledge graphs, which enable LLMs to perform multi-hop, cross-domain reasoning with improved factuality. Next, I will describe multi-task LLM pre-training techniques, including mixture-of-experts and instruction tuning, that enhance scientific information retrieval tasks such as patient-to-article retrieval and paper-reviewer matching by better aligning LLMs with domain-specific knowledge. Finally, I will present approaches for long-horizon reasoning and peer review, including long-chain-of-thought distillation for synthesizing deep research trajectories and tool-integrated agentic frameworks for generating evidence-grounded, high-quality reviews. Together, these methods demonstrate the potential of structure-aware LLMs to accelerate literature discovery, support research planning, and improve the quality of scientific feedback. Bio: Yu Zhang is an Assistant Professor in the Department of Computer Science & Engineering at Texas A&M University. He received his Ph.D. from the University of Illinois at Urbana-Champaign under Prof. Jiawei Han, and his B.Sc. from Peking University. His research focuses on natural language processing and data mining with structured knowledge, as well as their applications in scientific literature analysis. Yu is the ACM SIGKDD Dissertation Award Runner-Up, recipient of the UIUC Dissertation Completion Fellowship, and has received Best Paper at WSDM 2025 and Best Poster Honorable Mention at WWW 2018. He has published over 50 papers in leading conferences and journals, delivered 9 conference tutorials, and served as (Senior) Area Chair for venues including ACL, EMNLP, NeurIPS, and KDD. Thanks, Xuan -------- Assistant Professor Department of Computer Science Sanghani Center for AI and Data Analytics Virginia Tech