Dimensions overall score 8.0

GitHub Code Pulse

No public code linked for this paper yet.

• IntelliSA achieves an F1 score of 83% for IaC security smell detection, outperforming baseline static analyzers and LLMs by 7-42%.result 90%
• IntelliSA's student model is over 500x smaller than the LLM teacher model used for knowledge distillation.technical 90%
• IntelliSA can detect 60% of security smells while inspecting less than 2% of the codebase.result 90%
• Rule-based detection alone for IaC security smells tends to over-approximate and produce excessive false positives.limitation 85%
• Reliance on LLM APIs for security smell detection introduces high cost, latency, data governance concerns, and limits reproducibility and offline deployment.limitation 85%
• IntelliSA integrates symbolic rules with neural inference, using rules for over-approximation and neural inference for false positive filtering.method 90%
• The evaluation used a human-labeled dataset including 241 security smells across 11,814 lines of real-world IaC code.method 90%
• A single misconfiguration in IaC scripts can propagate widely, leading to severe system downtime and security risks.technical 80%

Competitive landscape

Competitor map is still being generated for this paper. Enable generation or check back soon.

Startup potential card