Measuring Diversity in Synthetic Datasets

Yuchang Zhu, Huizhe Zhang, Bingzhe Wu +5

Large language models (LLMs) are widely adopted to generate synthetic datasets for various natural language processing (NLP) tasks, such as text classification and summarization. H…