Monomoy is a tool for helping users familiarize themselves with and evaluate the behavior of machine learning models. It works by eliciting explanations of the model's behavior from the user and comparing them to explanations generated from the model.