In this homework, you will improve your understanding of how different models can behave very differently given the same prompt. You will also explore few-shot learning for a difficult natural language inference task. You will then experiment with solving a classification task using few-shot prompting techniques, and you will implement a method to mitigate label bias.