Cross-Domain Robustness of Llama3, Codestral, and DeepSeek-R1 in Vulnerability Classification

This report synthesises findings from 8 peer-reviewed papers addressing the following research question: What is the cross-domain robustness of Llama3, Codestral, and Deepseek R1 in classifying vulnerabilities when evaluated on both the original Big-Vul dataset and a synthetically generated variant,. 9 claims were extracted from source literature; 9 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 9.0/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: What is the cross-domain robustness of Llama3, Codestral, and Deepseek R1 in classifying vulnerabilities when evaluated on both the original Big-Vul dataset and a synthetically generated variant, measured by accuracy and precision-recall trade-offs?

Autonomous literature synthesis. Automated review score: 9.0/10. Full text and citation available at Assignee Research.