Skip to content
Review Open access

A Review of Source Code Vulnerability Detection Methods Based on Large Language Models: Models, Techniques, and Evaluation

Oct 2026 · Computers and artificial intelligence · 0 citations · 30 references

TL;DR

An application approach centered on data governance, strict data splitting, uncertainty calibration, and SAST-LLM-human collaboration is discussed, limited to source-code-level vulnerability detection, localization, and classification.

Abstract

In source code vulnerability detection, large language models (LLMs) are no longer used only as substitutes for traditional fixed classifiers. With prompt design, parameter fine-tuning, and program-context information, they can judge whether code contains vulnerabilities and can also provide suspicious locations and explanatory evidence. This paper takes lightweight code Transformers and large-scale language models as two main lines of discussion. It reviews methods for input construction, task adaptation, the use of program-analysis information, and multi-model collaboration, and compares them in terms of datasets, evaluation metrics, and deployment conditions. Existing studies show that code pretraining, prompt engineering, supervised fine-tuning, and structural-information enhancement can improve detection or explanation results on specific benchmarks. However, such improvements are easily affected by duplicate samples, label noise, random splits, temporal leakage, and changes in model versions, which may lead to overestimated performance. Therefore, model evaluation should not rely on a single metric alone. It should also report the data source, detection granularity, temporal or project-level isolation, false-positive cost, and whether generated explanations can be verified by program evidence. On this basis, this paper further discusses an application approach centered on data governance, strict data splitting, uncertainty calibration, and SAST-LLM-human collaboration. The discussion is limited to source-code-level vulnerability detection, localization, and classification, and does not cover network intrusion detection, malware analysis, smart-contract-specific detection, vulnerability generation, or automatic repair.

Read PDF

Similar papers

Preprint Oct 2026

Enhancing LLM-Based Vulnerability Detection via Representation-Aware Logical Ensemble

Automated vulnerability detection in C source code is a long-standing challenge, as traditional static analysis tools often trade precision for scalability and struggle with complex vulnerability patterns. Recently, large language models (LLMs) have shown promising results; yet, most approaches rely on a single code re...

Francesco C. Grasso, Pietro Liguori, R. Natella et al. · 0 citations
Open access 2026

An Uncertainty-Aware Adaptive Hybrid Framework for Robust Source Code Vulnerability Detection

: Source code vulnerability detection is important for improving software security and reliability. Pre-trained code representation models, such as CodeBERT and GraphCodeBERT, have improved vulnerability detection by learning semantic and structural characteristics from source code. However, existing approaches often r...

Tuan Nguyen Kim, Chieu Ta Quang, Nin Ho Le Viet · 0 citations
Preprint Oct 2026

On the Reliability of LLM-Based Vulnerability Patching Benchmarks

These results show that benchmark scores are highly sensitive to evaluation design, and practical guidelines for more rigorous, reliable, and reproducible evaluation are provided.

Dang K. Le, Wen-Xuan Shi, Xin-Yu Xing · 0 citations
Review Open access Oct 2026

Chakraview: A Hybrid Static-Analysis and Local-Llm Prototype for Vulnerability Detection and Mitigation Guidance in Open-Source Software

Open-source software benefits from rapid, distributed development, but reviewing it for security weaknesses is labour-intensive, and developers need findings that are explained and actionable. We present Chakraview, a prototype that combines a rule-based static stream with a locally hosted large language model (DeepSee...

Amit Rushikesh Tayade, Ajay Vinod Dawkhare, Aryan Mandar Wable, Mayur Shashikant Thakare, Sakshi Sopan Bhor · 0 citations

Metrics Failure in LLM-Based Code Vulnerability Repair: An Empirical Study and a Change-Aware Screen

It is argued that compile rate is a scientifically unreliable metric for single-function vulnerability repair, and this is supported with five controlled experiments over 203 vulnerable functions from Big-Vul, three open-source code LLMs (350M to 6.7B parameters), and three prompting strategies.

Om Nepal, Sushant Aryal, Oluseyi Olukola et al. · 0 citations
Open access 2026

Fine-Tuning UniXcoder for Code Smell Detection in Java Projects

This study investigates the application of UniXcoder, a pre-trained transformer model for source code, to classify Java source code methods across multiple projects as smell or clean, with a particular focus on the Switch Statements smell, and confirms that pre-trained transformer models, particularly UniXcoder, are ca...

Hanson Prihantoro Putro, Umi Laili Yuhana, E. M. Yuniarno et al. · 0 citations

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.