Large language model is very useful when:
We consider the model’s test is good, but its agency hasn’t reached the same level. The gap can be closed by leveraging the external rollout.