Retrieval-Enhanced Machine Learning

Retrieval-Enhanced Machine Learning (REML) explores how machine learning models can be augmented with retrieval components to improve performance, interpretability, and scalability. This framework treats information access systems as supporting both human users and task-driven machines, applying core principles of indexing, representation, retrieval, and ranking to enhance model capabilities. REML has applications across diverse domains including natural language processing, computer vision, time series prediction, and computational biology. Recent work explores fairness considerations in retrieval-augmented generation (RAG) systems, ensuring equitable exposure of sources and balanced attribution in generated content.