Supplementary data for “Benchmarking Local Vision-Language Models and Knowledge-Graph-Assisted Differential Diagnosis of Plant Diseases”

<p dir="ltr">Supplementary datasets for “Benchmarking Local Vision-Language Models and Knowledge-Graph-Assisted Differential Diagnosis of Plant Diseases.”</p><p dir="ltr"><b>SupplementaryData1.xlsx:</b> Qwen3.5-35B-A3B direct-classification predictions, matched retrieval results, and inference settings for 5,373 internal-evaluation images.</p><p dir="ltr"><b>SupplementaryData2.xlsx:</b> Symptom descriptions, ranked disease candidates, similarity scores, and prediction outcomes for the same 5,373 images evaluated using six vision-language models.</p><p dir="ltr"><b>SupplementaryData3.xlsx:</b> Neo4j knowledge-graph records used in AI Plant Doctor version 2, including hosts, diseases, symptoms, pathogens, pesticide information, reference-image cases, and model outputs.</p><p dir="ltr">The files contain derived outputs and structured records, not original images.

Raw disease labels are retained. Results represent internal evaluation rather than independent field validation.</p>