Objective:This test explores ChatGPT’s zero-shot generalization abilities. In other words, can it perform a task it has never explicitly been trained or prompted to do, without instructions?
Specifically, we tested whether the model could classify a ...
