What is few-shot prompting in an LLM, and how does it differ from fine-tuning?
Instruction: Explain using a small number of task examples in the inference prompt. Contrast this with updating model weights from a small training set, and describe how you would evaluate example quality.
Context: Distinguish in-context few-shot prompting from the broader family of few-shot learning methods.
Updated
Official answer available
Read the opening below, then unlock the full answer and practical guidance.
Few-shot prompting puts a small number of input/output demonstrations in the inference context. The model conditions its next response on those examples; ordinary prompting does not update its weights...