Prompt Engineering: A Comparative Study of Prompting Techniques in AI Language Models
Jothan Almeida · 2025
Prompt engineering plays a pivotal role in leveraging the full potential of large language models. This research paper focuses specifically on a comparative analysis of zero-shot and multi-shot prompting techniques, two widely used approaches in prompt engineering. While zero-shot prompting relies solely on task instructions to guide the model's output, multi-shot prompting incorporates examples within the prompt to provide additional context and improve task performance. As artificial intelligence and machine learning continue to advance, understanding how these distinct prompting strategies influence model behavior and output quality has become increasingly critical. Through empirical evaluations and performance metrics such as accuracy, precision, and recall, this paper examines the strengths and limitations of both techniques across various NLP tasks. By shedding light on their unique characteristics, this study aims to provide actionable insights for researchers and practitioners seeking to optimize LLM performance through effective prompt design. In addition to evaluating their practical applications, the findings of this paper underscore the importance of tailoring prompting strategies to the complexity and context of specific tasks. The analysis presented here contributes to a deeper understanding of how prompt structure impacts the capabilities of LLMs, laying a foundation for further innovations in prompt engineering.