Open soyisou opened 3 weeks ago
Hello, how are the simple propositions corresponding to the compound propositions in the training set obtained? Are they generated manually or by GPT?
We observe the data structure and apply rules manually to gain the simple propositions. With the emergence of large language models, we will be able to do better with large language models.
Hello, how are the simple propositions corresponding to the compound propositions in the training set obtained? Are they generated manually or by GPT?