Abstract: This presentation explores the prospect of conceptualizing college degree progress as a continuous framework of fine-grained mastery learning for transfers and others taking non-normative paths through higher education. First, I will discuss empirical studies demonstrating the efficacy of Large Language Models (LLMs) in generating psychometrically sound questions and pedagogically effective hints within Intelligent Tutoring Systems, building upon the open-source Open Adaptive tutor project being piloted at Cal. Second, I will extend this micro-level learning framework to macro-level institutional progress by examining stakeholder-informed AI methodologies for course articulation. Specifically, I will detail a supervised alignment method implemented within the SUNY system that mitigates institutional catalog biases, achieving a 5.5-fold improvement in accuracy and significantly expanding credit mobility. By leveraging representational learning techniques that can map learning objectives across contexts, the presentation will end with a questioning of how more flexible pathways can be constructed while retaining the value of a full-package University experience.
