AI-MI Seminar Series: Heng Ji (University of Illinois Urbana-Champaign & NSF MMLI)

Title: Agentic Modular Chemical Language Model (mCLM-A) for Molecule Discovery

Speaker: Heng Ji, University of Illinois Urbana-Champaign & Molecule Maker Lab Institute (NSF-MMLI)

Abstract: Scientific discovery is one of humanity’s greatest engines of progress, yet the process itself remains painfully slow, expensive, and largely artisanal. Foundation models have transformed AI into a powerful engine for acquiring, organizing, and reasoning over human knowledge. Yet today’s generative AI systems remain largely optimized for learning and reproducing patterns from what is already known, rather than discovering what is not yet known.
Human scientists operate fundamentally differently. They integrate heterogeneous evidence to identify unexplained phenomena, formulate competing hypotheses, reason about underlying mechanisms, design experiments, and revise their understanding through interaction with the physical world. Scientific intelligence emerges not simply from knowledge or reasoning, but from a continuous cycle of observation, hypothesis generation, verification, and knowledge updating.
In this talk, I argue that scientific discovery should inspire not only what AI is used for, but how AI itself is built. I will introduce a new paradigm of Science-Inspired Agentic AI that reimagines large language models as executable scientific agents rather than conversational assistants. The framework mirrors the scientific method through three tightly coupled capabilities: (1) Observe—acquire, represent, connect, and continuously update heterogeneous multimodal scientific knowledge; (2) Think—perform long-horizon, mechanism-aware reasoning to discover hidden connections and generate, challenge, and prioritize competing hypotheses; and (3) Propose and Verify—translate hypotheses into executable experiments, test them in computational and physical environments, and use experimental feedback to refine both hypotheses and knowledge.
Building on these principles, I will present an agentic AI framework that turns scientific reasoning into an executable, verification-centric discovery loop. Through applications in drug discovery and material discovery, I will demonstrate how this paradigm can move AI beyond retrieving and recombining existing knowledge toward discovering new hypotheses, testing them against reality, and ultimately generating new scientific knowledge.

Bio: Heng Ji is a Professor of Computer Science at Siebel School of Computing and Data Science of University of Illinois Urbana-Champaign. She is the Founding Director of Amazon-Illinois Center on AI for Interactive Conversational Experiences (AICE), and the Founding Director of CapitalOne-Illinois Center on AI Safety and Knowledge Systems (ASKS). She is a faculty member affiliated with Coordinated Science Laboratory, and Carl R. Woese Institute for Genomic Biology of University of Illinois Urbana-Champaign. She is named as an Association for Computational Linguistics (ACL) Fellow in 2025 for her “significant contributions to information extraction, multimodal and multilingual knowledge extraction and AI for science.” She received her Ph.D. in Computer Science from New York University under the supervision of the ACL Lifetime Achievement Awardee Prof. Ralph Grishman, and B.A. and M.A. in Computational Linguistics from Tsinghua University. The other awards she received include Outstanding Paper Award at ACL2024, two Outstanding Paper Awards at NAACL2024, “Young Scientist” by the World Laureates Association in 2023 and 2024, “Young Scientist” and a member of the Global Future Council on the Future of Computing by the World Economic Forum in 2016 and 2017, “Outstanding Advisor of the Year” by UIUC Engineering Council in 2026, Amazon Scholar 2020-2025, “Women Leaders of Conversational AI” (Class of 2023) by Project Voice, “AI’s 10 to Watch” Award by IEEE Intelligent Systems in 2013, NSF CAREER award in 2009, PACLIC2012 Best paper runner-up, “Best of ICDM2013” paper award, “Best of SDM2013” paper award, ACL2018 Best Demo paper nomination, ACL2020 Best Demo Paper Award, NAACL2021 Best Demo Paper Award, Google Research Award in 2009 and 2014, IBM Watson Faculty Award in 2012 and 2014 and Bosch Research Award in 2014-2018. She has given many keynote speeches at top conferences including IJCAI2025 and EMNLP2025. She served as the Program Committee Co-Chair of many conferences including NAACL-HLT2018 and AACL-IJCNLP2022. She was elected as the North American Chapter of the Association for Computational Linguistics (NAACL) secretary 2020-2023.

The talk will be livestreamed on the AI-MI YouTube channel.