TACAS Invited Speaker Guy Van den Broeck
Guy Van den Broeck is a Professor in the Computer Science Department at UCLA, where he directs the StarAI Lab. His research lies at the intersection of machine learning, knowledge representation, and reasoning. His work has been recognized with awards from leading conferences, including AAAI, UAI, KR, and OOPSLA. Guy is the recipient of an NSF CAREER award, a Sloan Fellowship, and the IJCAI-19 Computers and Thought Award.
Talk
Symbolic Reasoning in the Age of Large Language Models
Time: April 16, 9-10am
Room: TBA
Today, reasoning is often equated with large language models generating chains of thought. Historically, however, reasoning meant something very different: executing algorithms that manipulate symbols to perform logical or probabilistic deduction and derive definite answers to questions about knowledge. In this talk, I argue that such old-fashioned ideas are highly relevant to reasoning with large language models today. In particular, I will demonstrate that integrating formal symbolic reasoning algorithms directly into the architecture of language models yields state-of-the-art capabilities in controllable text generation, alignment, robot planning, and mathematical reasoning. These capabilities are built on top of tractable probabilistic circuit models that approximate the distribution of the language model’s future behavior, and allow for efficient formal reasoning on the GPU.