Year: 2025
Type: article
Abstract: Prompt engineering has emerged as a critical discipline in optimizing the performance of large language models (LLMs) across diverse domains.This paper provides a structured overview of key prompt engineering strategies identified through a review of recent literature, including Zero-Shot, Few-Shot, Chain-of-Thought, Tree-of-Thought, Self-Consistency, Directional Stimulus, and Heuristic and Ensemble Prompting.Each strategy is analyzed in terms of its conceptual foundation, practical applications, strengths, and limitations.The study highlights that effective prompt design depends not only on clarity, context, and structure but also on the strategic alignment between the prompting approach and the task's cognitive or computational demands.By comparing these strategies, the paper offers a comprehensive framework for selecting and combining prompt techniques to enhance reasoning, reliability, and adaptability in LLM-driven applications.
Author: Carlos J. Costa
Institution: University of Lisbon
Language: English
Topic: Topic Modeling
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
OpenAlex ID: W4416072221