multi hop questions
technique judged 86% likely a named technique or architecture
See how it has spread across media tiers
Mentions over time
Table view
| Week | Articles | Sources |
|---|---|---|
| 2024-W26 | 1 | 1 |
| 2026-W09 | 1 | 1 |
| 2026-W11 | 4 | 1 |
| 2026-W24 | 1 | 1 |
| 2026-W29 | 1 | 1 |
| 2026-W37 | 1 | 1 |
| 2026-W39 | 2 | 2 |
| 2026-W40 | 1 | 1 |
In their words
Cover easy lookups, multi-hop questions, ambiguous queries, and questions that have no answer in your corpus, since a good system should say so.
Repeated multi-hop questions are a strong signal.
Contextual retrieval provides more relevant answers by pulling connected information even when wording differs, reducing hallucinations through better grounding, making multi-hop questions queryable, enabling clear inspection/debugging of decision traces, and lowering token costs.
Multi-hop questions (‘which accounts share a beneficial owner?’) need a graph to answer.
GraphRAG outperforms flat retrieval because a knowledge graph captures the relationships that flat retrieval misses; multi-hop questions, such as “which accounts share a beneficial owner?” need a graph to navigate multiple hops out.
MosaicLeaks proposes a new deep-research task with multi-hop questions that interleave public and private information.
Common failure modes: - Dataset covers only “easy” queries that work well but misses edge cases - Dataset was created from the same documents used in the retrieval index (artificially high recall) - Dataset questions are too similar (testing the same retrieval path repeatedly) Mitigate by: reviewing dataset diversity regularly, adding real user queries as they are discovered, and testing adversari
The graph results provide the structured relationship chains that multi-hop questions require.