Few-Shot

Few-Shot (few-shot learning) is a learning paradigm in machine learning and artificial intelligence in which a model completes task learning or inference with only a very small number of labeled examples, typically ranging from one to a few dozen. Its core characteristics include using a small number of examples to guide rapid model adaptation to new tasks, and embedding a few examples in prompts to elicit in-context learning capabilities in large language models. It is contrasted with zero-shot learning and full supervised learning.