By Grace Stanley
Yoav Artzi, associate professor of computer science at Cornell Tech and the Cornell Ann S. Bowers College of Computing and Information Science, has received the Test of Time Award as part of the inaugural Transactions of the Association for Computational Linguistics (TACL) Paper Awards. The award was announced at the ACL 2025 conference, a leading international event for research in natural language processing, which took place in Vienna, Austria, from July 27 to August 1.
The Test of Time Award recognizes research that continues to shape the field years after its publication. Artzi was honored for his 2013 paper, Weakly Supervised Learning of Semantic Parsers for Mapping Instructions to Actions, alongside co-author Luke Zettlemoyer, professor at the University of Washington.
The paper’s approach has become foundational in building AI systems that interact naturally with people, influencing technologies ranging from voice assistants to robotics. The award recognizes the paper’s lasting impact on the field of natural language understanding and its role in advancing instruction-following capabilities in artificial intelligence.
The paper introduced a method for teaching computers to understand and follow human instructions without needing detailed, manual explanations. Instead, the system learned by watching examples unfold — a form of “weak supervision” — and tracking whether its actions led to successful outcomes, like reaching a destination or completing a task.
Overall, Artzi’s paper made a significant stride in building AI that can learn from interaction and context. At Cornell Tech, Artzi’s work combines language modeling and machine learning to build AI systems that can learn from context and interact intuitively with people. His group’s research often extends into robotics, computer vision, and cognitive science.
Grace Stanley is the staff writer-editor at Cornell Tech.
