An Expert-Curated Benchmark for Frontier and Local LLMs Querying a Biomedical Knowledge Graph
Johannes Fenoude Pancrace Medagbe, Munyoki Kilyungi, Arun Isaac, Alexander Kabua, Felix Lisso, Sammy Wambua, Robert L. Williams, Pjotr Prins
The paper benchmarks frontier and local LLMs on a biomedical knowledge graph using expert-curated questions and a new evaluation metric called psi.
bioRxiv · Oct 11, 2026 · PreprintBioinformatics