Showing 1–1 of 1 results for author: Brennan, H M

Trustworthiness of Children Stories Generated by Large Language Models

Authors: Prabin Bhandari, Hannah Marie Brennan

Abstract: Large Language Models (LLMs) have shown a tremendous capacity for generating literary text. However, their effectiveness in generating children's stories has yet to be thoroughly examined. In this study, we evaluate the trustworthiness of children's stories generated by LLMs using various measures, and we compare and contrast our results with both old and new children's stories to better assess th… ▽ More Large Language Models (LLMs) have shown a tremendous capacity for generating literary text. However, their effectiveness in generating children's stories has yet to be thoroughly examined. In this study, we evaluate the trustworthiness of children's stories generated by LLMs using various measures, and we compare and contrast our results with both old and new children's stories to better assess their significance. Our findings suggest that LLMs still struggle to generate children's stories at the level of quality and nuance found in actual stories △ Less

Submitted 25 July, 2023; originally announced August 2023.

Comments: 10 pages, 4 figures. To be published in 16th International Natural Language Generation Conference