Retrieval-Augmented Few-Shot Prompting Versus Fine-Tuning for Code Vulnerability Detection
This addresses the challenge of selecting effective in-context examples for LLMs in security-relevant code analysis, offering a practical alternative to fine-tuning, though it is incremental in improving few-shot methods.
The paper tackles the problem of improving few-shot prompting for code vulnerability detection by comparing retrieval-augmented prompting with standard few-shot prompting and fine-tuning, finding that retrieval-augmented prompting achieves an F1 score of 74.05% and partial match accuracy of 83.90% at 20 shots, outperforming fine-tuned Gemini but not CodeBERT.
Few-shot prompting has emerged as a practical alternative to fine-tuning for leveraging the capabilities of large language models (LLMs) in specialized tasks. However, its effectiveness depends heavily on the selection and quality of in-context examples, particularly in complex domains. In this work, we examine retrieval-augmented prompting as a strategy to improve few-shot performance in code vulnerability detection, where the goal is to identify one or more security-relevant weaknesses present in a given code snippet from a predefined set of vulnerability categories. We perform a systematic evaluation using the Gemini-1.5-Flash model across three approaches: (1) standard few-shot prompting with randomly selected examples, (2) retrieval-augmented prompting using semantically similar examples, and (3) retrieval-based labeling, which assigns labels based on retrieved examples without model inference. Our results show that retrieval-augmented prompting consistently outperforms the other prompting strategies. At 20 shots, it achieves an F1 score of 74.05% and a partial match accuracy of 83.90%. We further compare this approach against zero-shot prompting and several fine-tuned models, including Gemini-1.5-Flash and smaller open-source models such as DistilBERT, DistilGPT2, and CodeBERT. Retrieval-augmented prompting outperforms both zero-shot (F1 score: 36.35%, partial match accuracy: 20.30%) and fine-tuned Gemini (F1 score: 59.31%, partial match accuracy: 53.10%), while avoiding the training time and cost associated with model fine-tuning. On the other hand, fine-tuning CodeBERT yields higher performance (F1 score: 91.22%, partial match accuracy: 91.30%) but requires additional training, maintenance effort, and resources.