Retrieval-Augmented Few-Shot Prompting Versus Fine-Tuning for Code Vulnerability Detection

This article likely compares two different approaches for using Large Language Models (LLMs) to detect vulnerabilities in code. It contrasts retrieval-augmented few-shot prompting, which uses external knowledge to improve prompts, with fine-tuning, which adapts the LLM to a specific task. The research likely evaluates the performance of each method.