The Future of Code Analysis: Leveraging ChatGPT and CodeBert for Enhanced Security Insights

Go to Home » cyber

Abstract:

LLMs can be utilized in various tasks related to code analysis, such as code review, vulnerabilities assessment, and more. Despite their potential, the true strengths and limitations of integrating LLMs into code analysis remain uncertain. This study delves into the capabilities of LLMs in security-oriented program analysis, exploring insights from both attackers and security analysts. The analysis centers on two prominent LLMs, namely ChatGPT and CodeBert.

For further reading, you can access the full paper here.

If you are interested in similar articles, you may also want to check out our related topics:

• Code
• Analysis
• Paper
• Capabilities
• Both
• Analysts