Research Experiences for Undergraduates (REU), NHERI 2023: Comparing methods to Reduce Hallucinations in Large Language Models: A study of Cosine Similarity, Fine-Tuning and Zero-shot learning.

Large Language Models have revolutionized natural language processing and artificial intelligence, showcasing remarkable capabilities in generating human-like text and assisting with various task. However, it risks generating plausible but incorrect information (hallucinations) that lack factual grounding. This study presents a comprehensive comparison of three method for reducing hallucinations in LLMs: Cosine similarity, Fine-Tuning and Zero-shot learning. This project can be reused by both researchers and professionals in the field of Artificial intelligence and it is unique because it does a comparative analysis of methods to mitigating hallucinations. The intended audience of this research paper are professionals and students who working on Large Language Models, Natural Language Processing and Machine Learning.