ReGAIN: Retrieval-Grounded AI Framework for Network Traffic Analysis

Modern networks generate vast, heterogeneous traffic that must be continuously analyzed for security and performance. Traditional network traffic analysis systems, whether rule-based or machine learning-driven, often suffer from high false positives and lack interpretability, limiting analyst...

RESCUE: Retrieval Augmented Secure Code Generation

Despite recent advances, Large Language Models LLMs still generate vulnerable code. Retrieval-Augmented Generation RAG has the potential to enhance LLMs for secure code generation by incorporating external security knowledge. However, the conventional RAG design struggles with the noise of raw...