New Symbolic Reasoning: DecisionFlow - Advancing Large Language Models as Principled Decision Makers

Learn about DecisionFlow, a new symbolic reasoning AI model introduced for critical decision support that makes Chain of Thought reasoning 30% more powerful. This 35-minute presentation from University of Illinois Urbana-Champaign researchers explains how DecisionFlow advances large language models as principled decision makers through symbolic utility reasoning. The talk covers the research findings from authors Xiusi Chen, Shanyong Wang, Cheng Qian, Hongru Wang, Peixuan Han, and Heng Ji, providing insights into AI reasoning, science and technology, prompt design, and coding applications.