We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.CL

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo

Computer Science > Computation and Language

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

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.
Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
DOI: 10.1145/3626253.3635608
Cite as: arXiv:2404.09366 [cs.CL]
  (or arXiv:2404.09366v1 [cs.CL] for this version)

Submission history

From: Jaromir Savelka [view email]
[v1] Sun, 14 Apr 2024 21:38:50 GMT (515kb)

Link back to: arXiv, form interface, contact.