Abstract

Shared contextual understanding is fundamental to effective human communication. It conditions how meaning is interpreted when it is only partially specified. Although recent advances in conversational artificial intelligence (AI) have transformed how we access information and interact with machines, the key barriers that prevent current AI systems from functioning as human-centered collaborators are the lack of shared contextual understanding between human and AI and the ability to effectively navigate under such conditions. In this thesis, I present my research for overcoming these challenges through a framework of context synchronization, the coordination of distinct systems such that they align and operate in harmony according to a shared contextual understanding. First, I investigate methods for macro context synchronization, contextualization that applies to a broad population, to adapt language models …

Date

March 13, 2026

Authors

Hyundong Justin Cho

Institution

University of Southern California