Title: From Papers to Physical Designs: Empowering LLM with Integrated Datasets
Abstract: In the realm of circuit logic and machine learning models (LLMs), the availability and quality of datasets play a pivotal role in advancing research and development. This talk explores the critical importance of datasets in driving circuit LLM, focusing on a diverse array of datasets ranging from academic papers to commercial device specifications and physical designs. We delve into specific techniques, such as converting images to netlists, web crawling for extracting paper content, and computer vision for extracting data from figures. By leveraging these datasets and techniques, we enhance the accuracy, efficiency, and applicability of LLM in circuit design and analysis.
Biography: Dr. Yongfu Li is a tenured Associate Professor at Shanghai Jiao Tong University (SJTU). He is also a Board of Governors (BoG), and R10 Member At Large of the IEEE Circuits and Systems (CAS) Society, an AdCom Members of the IEEE Biometrics Council, a member of the IEEE DataPort Steering Committee, Chair of the IEEE Data Competition Committee, the IEEE CASS Standard Activities Sub Division, and the IEEE Asia Pacific Conference on Circuits and Systems (APCCAS) Steering Committee. He has previously worked at the National University of Singapore and GLOBALFOUNDRIES before joining SJTU in 2019. Throughout his career, he has earned numerous academic, industrial, and IEEE awards, including the IEEE EAB Society/Council Professional Development Award (2023), IEEE MGA YP Achievement Award (2022), and IEEE YP Hall of Fame Award (2021).