RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM
Graph retrieval-augmented generation (GraphRAG) enhances large language models with structured knowledge, yet existing systems construct knowledge graphs in a single extraction pass, producing noisy entities and brittle retrieval. RAGU, an open-source modular GraphRAG engine, addresses this by separating extraction from consolidation: entities and relations pass through two-stage typed extraction, DBSCAN-backed deduplication, LLM summarization, and Leiden community detection. A key insight motivates a compact extractor: the skills an in-pipeline LLM needs - comprehension, extraction, reasoning
Lineage graph
Paper → model → repo connections mined from source citations (Tier-1 exact match).
Why these links exist
Every edge carries a method, confidence, and the source snippet that justified it — so bad links are debuggable.
- PossiblePossibly related (embedding) · 62%bibinprathap/VeritasGraph →
- PossiblePossibly related (embedding) · 60%yifanfeng97/Hyper-Extract →
- PossiblePossibly related (embedding) · 55%Benchmarked Graph-RAG vs. Graph-Free Multi-Hop RAG: The graph mostly bought us a massive rebuild bill, not accuracy. →
- PossiblePossibly related (embedding) · 54%Zipstack/unstract →
- PossiblePossibly related (embedding) · 25%Tencent/WeKnora →
“Possibly related via embedding similarity 0.55 (not asserted). Timestamp check: artifact slightly before paper (-11d).”
- LinkedLinked via arxiv author · 85%Mikhail Komarov →
“RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM”
- LinkedLinked via arxiv author · 85%Ivan Bondarenko →
“RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM”
- LinkedLinked via arxiv author · 85%Stanislav Shtuka →
“RAGU: A Multi-Step GraphRAG Engine with a Compact Domain-Adapted LLM”