Abstract

In this talk, I'll survey some of the key architectural and processing elements of LLMs, showing the development arc from the original transformer model to the more complicated Chain of Thought, Mixture of Experts, and agentic models that are in the news today. I'll conclude with a discussion on the nature of LLM intelligence and how it compares to cognitive models of human intelligence.

About the Speaker

Thomas Eskridge is a Professor of Computer Science and Human-centered Design at Florida Tech, and a researcher at the L3Harris Institute for Assured Information and the Florida Tech Research Institute.