Julie Carson-Berndsen

Improving Linguistic Diversity of Large Language Models with Possibility Exploration Fine-Tuning

2024 · arXiv (Cornell University)

While Large Language Models (LLMs) have made significant strides in replicating human-like abilities, there are concerns about a reduction in the linguistic diversity of their outputs. This results in the homogenization of viewpoints and perspectives, …