A Framework for Adaptive Knowledge-Augmented Mizo Large Language Models Using Retrieval-Augmented Generation and Continual Learning
This paper presents an adaptive knowledge-augmented framework for Mizo Large Language Models by combining Retrieval-Augmented Generation (RAG) with continual learning that harnesses semantic retrieval with dense embeddings and FAISS indexing, adaptive evidence re-ranking, parameter-efficient fine-tuning, and incremental knowledge updating to enhance factual accuracy and decrease hallucinations.