We explore a lightweight framework that adapts frozen large language models to analyze longitudinal clinical data. The approach integrates patient history and context within the language model space to generate accurate forecasts without model fine-tuning. Applied to neuropsychological assessments, it achieves accurate and reliable performance even with minimal training data, showing promise for early-stage Alzheimer's monitoring.

Artificial intelligence (AI) has been becoming a subject of many fields over time. Whether it will be capable of doing anything human beings can do or not is a big part of the arguments (Artificial General Intelligence -- AGI). On the other side, Artificial Narrow Intelligence (ANI) which comprehends some capabilities not only becomes more reliable in those capabilities but also expands its frame (i.e. Machine learning (ML) which is a vital tool for AI makes supervised learning possible. As a basic example, AI for self-driving cars needs huge data for better performance in determining which is a car and which is not, to keep optimum distance in traffic.