ModelRefs / Query Decomposition — AI Glossary
Query Decomposition — AI Glossary
Breaking a complex multi-hop query into simpler sub-queries, executing each against the retrieval index, and merging results.
Overview
Query decomposition (DECOMP, IRCoT, LlamaIndex SubQuestionQueryEngine) identifies atomic sub-questions from complex queries (e.g., 'Who founded the company that acquired X?'), retrieves evidence per sub-question, and synthesizes a final answer. Significantly improves multi-hop QA benchmarks (HotpotQA, MuSiQue).
Reference details
| Topic | rag |
|---|---|
| Last reviewed | 2026-06-24 |
Related terms
Commonly confused with
For questions that need more than one retrieval to answer — comparisons, or anything requiring a fact found in one place and a fact found in another. Rewriting and expansion both still issue a search for a single question; decomposition accepts that no single search can succeed, so it is the right tool exactly when the query is multi-hop and the wrong one when it is merely vague.
Continue your research
Use these connected ModelRefs sections to compare alternatives, inspect implementation paths, and review the evidence and governance boundaries relevant to Query Decomposition — AI Glossary.
Frequently asked questions
What is Query Decomposition?
Breaking a complex multi-hop query into simpler sub-queries, executing each against the retrieval index, and merging results.
What concepts are related to Query Decomposition?
Closely related concepts include query rewriting, multi document qa, advanced rag.