Node Classification on Pubmed LLM-embedded (test)
85.05AccuracyRGVT + MLP
Evaluation results
| Method | Links | |
|---|---|---|
| RGVT + MLPTraining Protocol=Fully inductive (Pretrained on OGBN-Arxiv), Feature Embedding=LLM (text-embedding-3-large), Predictor=MLP2025.12 | 85.05 | |
| RGVT + MLPFeatures=LLM-embedded, Evaluation Setting=Fully inductive, Pretraining=OGBN-Arxiv2025.12 | 85.05 | |
| MLPFeatures=LLM-embedded, Evaluation Setting=Dataset-specific training2025.12 | 84.75 | |
| UniMPTraining Protocol=Tuned and trained separately per dataset, Feature Embedding=LLM (text-embedding-3-large)2025.12 | 84.71 | |
| UniMPFeatures=LLM-embedded, Evaluation Setting=Dataset-specialized tuning2025.12 | 84.71 | |
| GPRGNNTraining Protocol=Tuned and trained separately per dataset, Feature Embedding=LLM (text-embedding-3-large)2025.12 | 84.59 | |
| GPRGNNFeatures=LLM-embedded, Evaluation Setting=Dataset-specialized tuning2025.12 | 84.59 | |
| GCNTraining Protocol=Tuned and trained separately per dataset, Feature Embedding=LLM (text-embedding-3-large)2025.12 | 83.46 | |
| GCNFeatures=LLM-embedded, Evaluation Setting=Dataset-specialized tuning2025.12 | 83.46 | |
| GraphAnyFeatures=LLM-embedded, Evaluation Setting=Fully inductive, Pretraining=OGBN-Arxiv2025.12 | 78.69 |