Table 9 - long-short matching subgroup - text classification Approach: Generate dataset consisting of n text classification tasks using the prompt from table 9 Generate the actual training dataset using the second prompt in table 9. Sample from the dataset of classification tasks.