rag-replug-2024-blackbox
IN premise — summaries/2026/08/24/wiki-Retrieval-augmented_generation-chunk-2.md
Created 2026-08-25T02:58:52+00:00
REPLUG (Shi et al., 2024) enables in-context RAG with black-box LLMs by prepending retrieved documents to the prompt without requiring model weight access.
Summary
You can plug retrieved context into any LLM's input prompt and get retrieval-augmented answers without ever touching the model's weights or having a hosted copy of it. This matters because it makes RAG available to anyone with an API key, removing the need for expensive self-hosted infrastructure or fine-tuning pipelines.