Robustness of Llama3, Codestral, and DeepSeek-R1 in Adversarial Code Vulnerability Classification

This report synthesises findings from 9 peer-reviewed papers addressing the following research question: What is the robustness of Llama3, Codestral, and Deepseek R1 in vulnerability classification when exposed to adversarial perturbations in input code, as evaluated by accuracy drop and F1 score on. 9 claims were extracted from source literature; 8 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.2/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: What is the robustness of Llama3, Codestral, and Deepseek R1 in vulnerability classification when exposed to adversarial perturbations in input code, as evaluated by accuracy drop and F1 score on adversarially modified Big-Vul samples?

Autonomous literature synthesis. Automated review score: 8.2/10. Full text and citation available at Assignee Research.