Understanding the Role of Temperature in Diverse Question Generation by GPT-4

Arav Agarwal; Karthik Mittal; Aidan Doyle; Pragnya Sridhar; Zipiao Wan; Jacob Arthur Doughty; Jaromir Savelka; Majd Sakr

doi:10.1145/3626253.3635608

Understanding the Role of Temperature in Diverse Question Generation by GPT-4

Computation and Language 2024-04-16 v1 Artificial Intelligence

Authors: Arav Agarwal , Karthik Mittal , Aidan Doyle , Pragnya Sridhar , Zipiao Wan , Jacob Arthur Doughty , Jaromir Savelka , Majd Sakr

View on arXiv ↗ PDF ↗ DOI ↗

Abstract

We conduct a preliminary study of the effect of GPT's temperature parameter on the diversity of GPT4-generated questions. We find that using higher temperature values leads to significantly higher diversity, with different temperatures exposing different types of similarity between generated sets of questions. We also demonstrate that diverse question generation is especially difficult for questions targeting lower levels of Bloom's Taxonomy.

Keywords

question generation GPT-4 temperature parameter diversity Bloom's Taxonomy

Cite

@article{arxiv.2404.09366,
  title  = {Understanding the Role of Temperature in Diverse Question Generation by GPT-4},
  author = {Arav Agarwal and Karthik Mittal and Aidan Doyle and Pragnya Sridhar and Zipiao Wan and Jacob Arthur Doughty and Jaromir Savelka and Majd Sakr},
  journal= {arXiv preprint arXiv:2404.09366},
  year   = {2024}
}

Understanding the Role of Temperature in Diverse Question Generation by GPT-4

Abstract

Keywords

Cite

Related papers