Node Classification on Cora Arxiv+Pubmed
16.43AccuracyLoReC(GraphGPT-cot)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LoReC(GraphGPT-cot)Decoding Strategy=LoReC, Base Prompting=cot, Backbone=Vicuna-7b2026.04 | 16.43 | 13.94 | |
| Vicuna-7b-v1.5Backbone=Vicuna-7b2026.04 | 14.89 | 12.13 | |
| GraphGPT-cotDecoding Strategy=Chain-of-Thought, Backbone=Vicuna-7b2026.04 | 14.59 | 12.87 | |
| LoReC(GraphGPT-std)Decoding Strategy=LoReC, Base Prompting=std, Backbone=Vicuna-7b2026.04 | 12.78 | 7.85 | |
| GraphGPT-stdDecoding Strategy=Standard, Backbone=Vicuna-7b2026.04 | 12.59 | 7.73 | |
| GKD2026.04 | 4.06 | 0.37 | |
| Baichuan-7bBackbone=Baichuan-7b2026.04 | 4.05 | 4.69 | |
| GCN2026.04 | 1.87 | 0.32 | |
| GLNN2026.04 | 1.82 | 0.92 | |
| GAT2026.04 | 1.61 | 0.57 |