Methods, Modeling, and Findings

This unit examines large language models and the research methods used to study them, guiding students through experimental design, modeling approaches, and presentation of results. Students will learn to build or configure simple LLM workflows, analyze outcomes, and communicate findings and limitations in a structured, reproducible way.

Learning Objectives

• Analyze the architecture, capabilities, and limitations of large language models and connect those characteristics to appropriate AI-for-Good applications
• Apply research methods to design and conduct experiments, including data collection, preprocessing, annotation practices, and selection of evaluation metrics for LLM projects
• Demonstrate basic modeling skills by constructing or configuring a reproducible LLM pipeline (preprocessing, fine-tuning or inference setup) and documenting the process
• Evaluate and communicate findings, limitations, and ethical considerations by producing a structured report that follows the documented book structure and informs stakeholders