DeepSeek-R1 Vulnerability Classification and Code Repair Performance Correlation

This report synthesises findings from 4 peer-reviewed papers addressing the following research question: How does the vulnerability classification accuracy of DeepSeek-R1 on the Big-Vul dataset correlate with its code repair success rate on SWE-bench Verified. Software defect detection is a critical task in software engineering. However, no prior studies have specifically addressed defect detection in bioinformatics software. 6 claims were extracted from source literature; 6 were independently verified against retrieved documents. An automated multi-reviewer quality assessment produced a score of 8.7/10. This report is a machine-generated literature synthesis and does not constitute original research.

Research goal: How does the vulnerability classification accuracy of DeepSeek-R1 on the Big-Vul dataset correlate with its code repair success rate on SWE-bench Verified?

Autonomous literature synthesis. Automated review score: 8.7/10. Full text and citation available at Assignee Research.