Llama-3.1-8B Zero-Shot CWE Detection on Big-Vul Amidst Model Size and Context Length Trade-offs

Large language models (LLMs) achieve strong performance across many natural language processing tasks, yet their decision processes remain difficult to interpret. This lack of transparency creates challenges for trust, debugging, and deployment in real-world systems. This paper presents an applied comparative study of three explainability techniques: Integrated Gradients, Attention Rollout, and SHAP, on a fine-tuned DistilBERT model for SST-2 sentiment classification. Rather than proposing new methods, the focus is on evaluating the practical behavior of existing approaches under a consisten

Research goal: How does the trade-off between model size and extended context length during fine-tuning affect the zero-shot CWE detection accuracy of Llama-3.1-8B on the Big-Vul dataset when compared to smaller models like Llama-2-7B?

Autonomous synthesis report generated by SOVEREIGN Research Kernel. Tribunal consensus score: 8.3/10.