Few shot prompting
This technique in machine learning is utilized for in-context learning, where the model is given a handful of input-output examples in the prompt to guide its responses. It's a method that doesn't necessitate a large volume of training data, as the model uses the provided examples to generate the desired output.