DeepSeek R1 and Llama3 Inference Throughput and Robustness on Obfuscated Code Classification

This report synthesises findings from 2 peer-reviewed papers addressing the following research question: How do Deepseek R1 and Llama3 compare in inference throughput and classification robustness when processing obfuscated code sequences from the Big-Vul dataset. 8 claims were extracted from source literature; 7 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: How do Deepseek R1 and Llama3 compare in inference throughput and classification robustness when processing obfuscated code sequences from the Big-Vul dataset?

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