Home/Browse/Graphs/Node Classification/Cora

Cora

Unknown

Citation network of scientific papers. 2708 nodes, 5429 edges, 7 classes. Classic GNN benchmark.

Benchmark Stats

Models21
Papers21
Metrics1

SOTA History

accuracy

accuracy

Higher is better

RankModelSourceScoreYearPaper
1TAPE + RevGAT

TAPE (LLM-to-LM Interpreter) + RevGAT backbone. ICLR 2024. He et al. Uses LLM-generated explanations as node features. Supervised split.

Community92.92024Source
2AuGLM (T5-large)

AuGLM with T5-large backbone. Text-output node classifier. Xu et al. 2024 "How to Make LMs Strong Node Classifiers?" Table 1.

Community91.512024Source
3ENGINE

ENGINE vector-output model. Result from AuGLM comparison table (Table 1) in Xu et al. 2024.

Community91.482024Source
4InstructGLM

InstructGLM text-output model. Result from AuGLM comparison table (Table 1) in Xu et al. 2024.

Community90.772024Source
5GLEM + RevGAT

GLEM (Graph-LM EM framework) + RevGAT backbone. From AuGLM comparison table (Table 1) in Xu et al. 2024.

Community88.562024Source
6GCNLLMEmb

GCN with LLM-generated embeddings, supervised setting. From comprehensive LLM-based node classification analysis, Feb 2025.

Community88.152025Source
7LLaGA (Mistral-7B)

LLaGA with Mistral-7B backbone, supervised setting. Xu et al. 2024 Table 6.

Community87.552024Source
8SDGAT

Sparse graphs-based Dynamic Attention Network. ~3% improvement over baselines on Cora. Published PMC Dec 2024.

Community85.292024Source
9GCN* (tuned)

GCN with proper hyperparameter tuning. Best model in NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2). Luo et al.

Community85.082024Source
10GAT* (tuned)

GAT with proper hyperparameter tuning. NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community84.642024Source
11SGFormer

SGFormer result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2). Wu et al.

Community84.52024Source
12GraphSAGE* (tuned)

GraphSAGE with proper hyperparameter tuning. NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community84.182024Source
13Polynormer

Polynormer result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community83.252024Source
14GOAT

GOAT (Graph Transformer) result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community83.182024Source
15GAT

Graph Attention Network. Velickovic et al., ICLR 2018.

Community832018Source
16GraphGPS

GraphGPS result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2). Rampasek et al. original model.

Community82.842024Source
17Exphormer

Exphormer result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community82.772024Source
18GraphSAGE

GraphSAGE result from NeurIPS 2024 "Classic GNNs are Strong Baselines" paper (Table 2). Luo et al.

Community82.682024Source
19NodeFormer

NodeFormer result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community82.22024Source
20NAGphormer

NAGphormer result from NeurIPS 2024 "Classic GNNs are Strong Baselines" (Table 2).

Community82.122024Source
21GCN

Graph Convolutional Network. Kipf & Welling, ICLR 2017. Standard semi-supervised split.

Community81.52017Source

Submit a Result