Senior Thesis
School:
School of Computer and Information Sciences
Category:
Data SciencePrimary
Project Overview
One Liner: Exploring whether fine-tuning language models can help them tell the same story from different moral perspectives without changing the main events.
Abstract
Large language models can generate coherent narratives but may struggle to represent different moral perspectives without introducing inconsistencies or relying on superficial linguistic changes. This project aims to investigate whether fine-tuning can improve a language model's ability to generate narratives that reflect specific moral frameworks while preserving the underlying events and overall coherence.
No video available.
Screenshots
0 image(s)No screenshots uploaded yet.