Knowledge graphs lift GPT-4o-mini multi-hop retrieval from 36% to 56% at 0.05s cost, Vector Institute reports
VectorInst · x · 2026-10-09
Vector Institute published a full technical writeup on knowledge-graph-enhanced RAG, arguing structured context delivers high accuracy without heavy compute.
- On complex multi-hop queries over an SEC 10-Q dataset, GPT-4o retrieval accuracy rose from 40% to 55%
- Cost-effective GPT-4o-mini jumped from 36% to 56%, adding only 0.05 seconds of latency — structural context lets smaller models match larger ones
- The post covers the full extraction pipeline, evaluation methodology, comparisons with GraphRAG and Cypher-based approaches, and open-source code
Related event: Knowledge Graph RAG Boosts Multi-Hop Retrieval Accuracy by 54%(3 posts)→
More from Research
- Formalized classification of finite simple groups may be near, says Irving — geoffreyirving · 2026-10-10
- Success-Guided Sampling: zero real-world data for dexterous robot hands — milesmacklin · 2026-10-10
- Google DeepMind launches AI co-clinician initiative for 'triadic care' in medicine — alan_karthi · 2026-10-10
- Anthropic's Geoffrey Irving: is the formalized classification of finite simple groups coming next week? — geoffreyirving · 2026-10-10
- Cell-rejuvenating therapy shows early vision improvement in 2 of 3 glaucoma patients — kimmonismus · 2026-10-10
- Terence Tao: The Era of "Big M" Mathematics and "Big P" Physics Has Arrived — CatAstro_Piyush · 2026-10-10