Abstract
Recently, enhancing large language models (LLM) with mathematical reasoning has become a vibrant area of research. In this talk, Professor Tam will explore programmatic methods that guide an LLM to produce source code as reasoning steps, which can then be assessed by an external interpreter. He will also share some findings on using Prolog, a logic programming language, to tackle mathematical problems.
Biography
Yik-Cheung (Wilson) Tam is a Professor of Practice in Computer Science at NYU Shanghai. From 2017 to 2020, he was a principal NLP scientist and manager at WeChat AI focusing on research and development for Xiaowei, a virtual personal assistant, and music recommendation. From 2015 to 2017, Wilson was a senior NLP scientist and manager at Microsoft (Suzhou) working on natural language understanding for Cortana, Microsoft's virtual personal assistant. He has six years of research experience in the United States, including with SRI international (formerly Stanford Research Institute), and Nuance communications Inc.