Prompt-Based Classification

Prompt-based classification enables rapid adaptation to new labeling tasks, especially when using large language models. The model treats classes as options described in natural language and makes its decision in that context. This offers strong benefits in low-data settings, rapid prototyping, and variable label spaces. However, stability, bias, and prompt sensitivity must be managed carefully.