AI Paper Atlas

Research map / Large language models / Retrieval and knowledge

Distill-SynthKG: Distilling Knowledge Graph Synthesis Workflow for Improved Coverage and Efficiency

Prafulla Kumar Choubey, Xin Su, Man Luo, XIANGYU PENG, Caiming Xiong, Tiep Le, Shachar Rosenman, Vasudev Lal, Phil L Mui, Ricky Ho, Phillip Howard, Chien-Sheng Wu

ICLR 2026 · Large language models › Retrieval and knowledge · approach: Knowledge graph reasoning with LLMs

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use multi-step SynthKG pipeline instead of single-step prompting with Llama-3-8bLlama-3-8b KG construction · MuSiQue KG coveragetriplet coverage: 32.09 -> 38.84Improved
use multi-step SynthKG pipeline instead of single-step prompting with Llama-3-70bLlama-3-70b KG construction · 2Wiki KG coveragetriplet coverage: 54.10 -> 63.30Improved
distill SynthKG into single-step 8b model (D-SynthKG-8b) versus untrained Llama-3-8bLlama-3-8b · MuSiQue KG coveragetriplet coverage: 32.09 -> 46.90Improved
distilled 8b model versus its multi-step 70b teacher pipelineD-SynthKG-8b · 2Wiki KG coveragetriplet coverage: 63.30 -> 58.27Hurt
use Graph+LLM retriever on D-SynthKG-8b KG versus the Llama-3-8b KGGraph+LLM retrieval · MuSiQue retrievalHits@2: 31.33 -> 53.35Improved
use Graph+LLM on D-SynthKG-8b KG versus Llama-3-8b KG for multi-hop QAGraph+LLM QA · Multi-hop QA averageEM: 0.281 -> 0.433Improved
add LLM re-ranking on top of dense retrieval (Dense+LLM)Dense retriever · Multi-hop QA averageEM: 0.363 -> 0.394Improved
use proposition-entity Graph+LLM retriever instead of dense retrievalretrieval · MuSiQue, 2Wiki, HotpotQA retrievalHits@2: +12.75 points over dense, +1.67 over Dense+LLMImproved
retrieve propositions instead of triples (ablation)KG retrieval source · multi-hop QA retrievalHits@10: +0.89 Hits@10Improved
use full graph-based retrieval instead of propositions only (ablation)KG retrieval source · multi-hop QA retrievalHits@10: +2.50Improved
add propositions to retrieved chunks as QA context (ablation)RAG context · multi-hop QAEM: +2 EMImproved
add 2-hop paths to retrieved chunks as QA context (ablation)RAG context · multi-hop QAEM: +2.7 EMImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy