If I wanted to build a rag around a generative transformer, how does the retrieved vector plug into the transformer? Is it cross attention from an encoder or a

If I wanted to build a rag around a generative transformer, how does the retrieved vector plug into the transformer? Is it cross attention from an encoder or a prefix on the decoder, or sonething else? Does it plug in directly as the retrieved vector or is it somehow re-embedded?