Regular biography
James Thorne is a faculty member in the Department of Computer Science at KAIST. His research is associated with the Kim Jaechul Graduate School of AI. Thorne's work involves industry classification and factor models, serving as foundational intelligence for capital markets. He is also the CTO at Theia Insights. His academic profile includes contributions to NLP and computational linguistics, with a focus on parsing and incremental parsing. Thorne utilizes tools such as Overleaf and Docker for academic and research work. His website is https://jamesthorne.com/, and his email is thorne@kaist.ac.kr.
Scholar-generated biography
James Thorne is a researcher specializing in Natural Language Processing, with a focus on fact extraction, verification, and language model evaluation. His work includes developing large-scale datasets such as FEVER and KILT, which are used to benchmark knowledge-intensive language tasks. Thorne has also contributed to automated fact-checking methodologies and the creation of frameworks for verifying numerical claims. His research emphasizes the integration of reasoning and alignment skills in language models, as seen in projects like Prometheus and Flask. Additionally, he has explored hate speech detection and multi-task learning for domain transfer.