About the results #5
Description
The best performing model introduced in Chapter 4 is GPT-4o, which achieved 78.8%. But in Table 2, the two results of GPT-4o are 83.57% and 80.96%, respectively. In my understanding, there should be one that belongs to the default setting, but both are different from 78.8%.