liveBook · Manning

Sufficient context

As explained in Retrieval Augmented Generation, The Foundational Ideas

Sufficient context is the condition in which the retrieved set collectively contains the information needed to answer the query.

A retrieval system can fail quietly when it supplies partial evidence: models answer incorrectly more often with insufficient context than with no context. section 14.5.2

Sufficient context means that the retrieved set collectively contains the information needed to answer the query. It differs from relevance: relevant passages may still fail to justify the answer. section 14.5.2

Use the metric to guide evaluation and response policy:

  • Compare retrieval configurations rather than treating an absolute score threshold as universal. section 14.5.2
  • Use sufficiency judgments to decide when the system should abstain. section 14.5.2
  • Check corpus state separately: a superseded policy can answer the question completely while remaining outdated. section 14.5.2

Why partial evidence is dangerous

Gemma answered incorrectly on 10.2% of questions without context and 66.1% with insufficient context. A prompted autorater classified sufficiency correctly at least 93% of the time; abstention based on its labels raised answer correctness by 2–10 points at the same coverage. section 14.5.2

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Defines sufficiency, distinguishes it from relevance, reports autorater results, and identifies corpus comparison as unresolved.
Examples
Uses Gemma error rates, superseded policy evidence, and prompted autorater abstention results.
What it assumes
Assumes retrieval and answer scoring; teaches why relevance alone cannot establish sufficient evidence.
What you can do after
Evaluate retrieval configurations and diagnose when a system should abstain instead of answering.
How it connects
Connects sufficient context to retrieval, relevance, faithfulness, answer evaluation, and abstention.

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Retrieval Augmented Generation, The Foundational Ideas. The section links point to the parts of the book it draws on.