Beyond the Parameters: A Technical Survey of Contextual Enrichment in Large Language Models: From In-Context Prompting to Causal Retrieval-Augmented Generation

Researchers published a comprehensive technical survey on Large Language Model augmentation strategies, examining methods from in-context learning to advanced Retrieval-Augmented Generation techniques. The study provides a unified framework for understanding how structured context at inference time can overcome LLMs' limitations of static knowledge and finite context windows.