Definition
Precision@K measures the proportion of relevant documents retrieved within the top K results of a RAG pipeline's retrieval stage. It evaluates the 'signal-to-noise' ratio of the context provided to the LLM, directly influencing the model's ability to generate accurate answers without processing distracting or irrelevant data.
Measures retrieval purity in a specific window, rather than total knowledge coverage (Recall) or model generation accuracy.
"A gold miner's pan: Precision@K is the ratio of actual gold nuggets to worthless silt found in a single scoop of K size."
Conceptual Overview
Precision@K measures the proportion of relevant documents retrieved within the top K results of a RAG pipeline's retrieval stage. It evaluates the 'signal-to-noise' ratio of the context provided to the LLM, directly influencing the model's ability to generate accurate answers without processing distracting or irrelevant data.
Disambiguation
Measures retrieval purity in a specific window, rather than total knowledge coverage (Recall) or model generation accuracy.
Visual Analog
A gold miner's pan: Precision@K is the ratio of actual gold nuggets to worthless silt found in a single scoop of K size.