Loading
Life sciences · Preprint
arXiv · September 10, 2026
Posted before peer review. The findings may change or fail to hold.
EXYGEN is a proof-of-concept framework for converting natural language to database queries over knowledge graphs using large language models and retrieval-augmented generation, without task-specific fine-tuning. The authors report an exact-match accuracy of 0.419 on the SciQA benchmark and describe a scalable sampling strategy to generate metadata from very large graphs. This is a systems contribution intended to reduce fine-tuning burden; it is not clinically validated and has not been peer reviewed.
Preprint. Knowledge graphs: SciQA benchmark; OpenCitations Meta; GESIS; ORKG. No human subjects.. Intervention: EXYGEN framework: text-to-SPARQL generation using LLMs with RAG incorporating VoID descriptions, ShEx schemas, and retrieved triple examples; predicate-coverage-aware parallel graph sampling for metadata generation from large KGs.. Compared with: Ablations of the RAG pipeline components; comparison of different LLM sizes; implicit comparison to fine-tuned baselines (not directly evaluated)..
Best configuration achieved exact match of 0.419 on execution results for text-to-SPARQL generation without LLM fine-tuning Lexical metrics such as F1 poorly predict query correctness Larger general-purpose LLMs outperform smaller code-specialized ones given sufficient context
Safety was not reported in the material analysed. Check the source before drawing any conclusion about harm.
The source did not state who this applies to in practice.
This is an unrefereed technical preprint describing a framework and computational methods for knowledge graph querying; it reports engineering results and benchmark performance but has not undergone peer review.
As stated by the source record.
Quoted from the source exactly as published.
Graded across the dimensions that decide whether you should act, each from what the source actually supports. There is no single score, and where a dimension was not assessed it says so.
We present EXYGEN (EXplore Your Graphs ENgine), a framework for knowledge graph (KG) understanding that enables conversational access to KGs at scale. We address two questions in sequence. First, how effectively can LLMs perform text-to-SPARQL generation given only automatically derived structured metadata and small graph samples, rather than task-specific fine-tuning? We integrate VoID descriptions and ShEx schemas into a retrieval-augmented generation (RAG) pipeline and ablate KG-derived context on the SciQA benchmark. Our best configuration -- combining ShEx schemas, retrieved triples, and example question-query pairs -- reaches an exact match of 0.419 on execution results without any LLM fine-tuning. We further find that lexical metrics such as F1 poorly predict query correctness, and that larger general-purpose LLMs can outperform smaller code-specialized ones once given sufficient context. Second, we ask how to generate the structured metadata that this method relies on from very large KGs, where KG metadata generation becomes computationally intractable. We introduce a predicate-coverage-aware parallel graph sampling strategy that preserves structural diversity while remaining computationally tractable. On OpenCitations Meta and GESIS, it retains high predicate coverage with minimal triple loss and reduces runtime by over 80x; on ORKG, sampling is not just faster but the only tractable path to obtain complete metadata. Together, these results show that structured schema context and lightweight prompting can substantially reduce reliance on fine-tuning for scalable conversational access to KGs, though closing the remaining gap to fully fine-tuned approaches will likely require reducing dependence on curated question-query exemplars -- whether through synthetic generation or an execution-feedback-driven approach -- and validating these findings beyond a single benchmark.
Taken from the source record, never inferred. Follow any of these and new work involving them reaches your briefing.