Other Graph Benchmarks
Amazon2M
Structure
1Graphs
2,449,029Nodes
61,859,140Edges
47Classes
100Node Feature Dim
Coverage
215Results
1Variants
22Papers
- Learning Feature Encoder with Synthetic Anomalies for Weakly Supervised Graph Anomaly Detection Learning Feature Encoder with Synthetic Anomalies for Weakly Supervised Graph Anomaly Detection
- Turning Tabular Foundation Models into Graph Foundation Models Turning Tabular Foundation Models into Graph Foundation Models
- From Nodes to Narratives: Explaining Graph Neural Networks with LLMs and Graph Context From Nodes to Narratives: Explaining Graph Neural Networks with LLMs and Graph Context
- NeurIPS 2025: Semi-supervised Graph Anomaly Detection via Robust Homophily Learning Semi-supervised Graph Anomaly Detection via Robust Homophily Learning NeurIPS 2025
- Language Models as Messengers: Enhancing Message Passing in Heterophilic Graph Learning Language Models as Messengers: Enhancing Message Passing in Heterophilic Graph Learning
- ICLR 2024: Partitioning Message Passing for Graph Fraud Detection Partitioning Message Passing for Graph Fraud Detection ICLR 2024
- TMLR 2024: SCNode: Spatial and Contextual Coordinates for Graph Representation Learning SCNode: Spatial and Contextual Coordinates for Graph Representation Learning TMLR 2024
- Graph Triple Attention Network: A Decoupled Perspective Graph Triple Attention Network: A Decoupled Perspective
- NTFormer: A Composite Node Tokenized Graph Transformer for Node Classification NTFormer: A Composite Node Tokenized Graph Transformer for Node Classification
- TopoBench: A Framework for Benchmarking Topological Deep Learning TopoBench: A Framework for Benchmarking Topological Deep Learning
- ICML 2024: Rethinking Independent Cross-Entropy Loss For Graph-Structured Data Rethinking Independent Cross-Entropy Loss For Graph-Structured Data ICML 2024
- NeurIPS 2023: GADBench: Revisiting and Benchmarking Supervised Graph Anomaly Detection GADBench: Revisiting and Benchmarking Supervised Graph Anomaly Detection NeurIPS 2023
- Tokenized Graph Transformer with Neighborhood Augmentation for Node Classification in Large Graphs Tokenized Graph Transformer with Neighborhood Augmentation for Node Classification in Large Graphs
- WWW 2023: Node-wise Diffusion for Scalable Graph Learning Node-wise Diffusion for Scalable Graph Learning WWW 2023
- AAAI 2022: MA-GCL: Model Augmentation Tricks for Graph Contrastive Learning MA-GCL: Model Augmentation Tricks for Graph Contrastive Learning AAAI 2022
- Geometry Contrastive Learning on Heterogeneous Graphs Geometry Contrastive Learning on Heterogeneous Graphs
- ICML 2022: Rethinking Graph Neural Networks for Anomaly Detection Rethinking Graph Neural Networks for Anomaly Detection ICML 2022
- GAP: Differentially Private Graph Neural Networks with Aggregation Perturbation GAP: Differentially Private Graph Neural Networks with Aggregation Perturbation
- WWW 2021: Lorentzian Graph Convolutional Networks Lorentzian Graph Convolutional Networks WWW 2021
- Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters Enhancing Graph Neural Network-based Fraud Detectors against Camouflaged Fraudsters
- ICLR 2019: GraphSAINT: Graph Sampling Based Inductive Learning Method GraphSAINT: Graph Sampling Based Inductive Learning Method ICLR 2019
- KDD 2019: Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks KDD 2019
Node classification on a roughly two-million-node co-purchase graph, predicting each product’s category. Evaluated by Accuracy.
Dataset details
Task
Large-scale Amazon co-purchase graph with ~2M nodes. Node classification (product category). Used to benchmark scalable GNNs and graph transformers. The primary catalog metric is Accuracy.
Data
Large-scale Amazon co-purchase graph with ~2M nodes. Node classification (product category). Used to benchmark scalable GNNs and graph transformers. The cataloged artifact reports 1 graphs.
Features
Input construction is dataset-specific and follows the cited release: program call graphs, 3D point coordinates, text-derived vectors, product attributes, or movie metadata. Added edges, neighborhood graphs, and feature normalization are preprocessing choices.
Splits and evaluation
The catalog records these protocols or variants: Standard split. Evaluation uses Accuracy (higher is better). Exact masks or folds must come from the cited release.
Quirks and pitfalls
These entries come from unrelated upstream benchmarks and have no shared protocol. Report the exact graph construction, split, feature pipeline, and label granularity; similarly named Film, Actor, IMDB, or Amazon artifacts are not interchangeable.
Sources
SOTA Progression
new SOTA
matches SOTA at that time
other proposed result
other baseline result
Leaderboard
| # | Model | Accuracy ↑ | AUC | AUPRC | AUROC | F1 | F1-macro | F1-micro | Fidelity | Macro-F1 | micro-F1 | Paper |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | MA-GCL | 0.8883
±
0.0030 | — | — | — | — | — | — | — | — | — | |
| 2 | CCA-SSG | 0.8842
±
0.0030 | — | — | — | — | — | — | — | — | — | |
| 3 | ARIEL | 0.8827
±
0.0020 | — | — | — | — | — | — | — | — | — | |
| 4 | BGRL | 0.8823
±
0.0030 | — | — | — | — | — | — | — | — | — | |
| 5 | GCA | 0.8785
±
0.0030 | — | — | — | — | — | — | — | — | — | |
| 6 | Base Model | 0.8765
±
0.0020 | — | — | — | — | — | — | — | — | — | |
| 7 | MVGRL | 0.8752
±
0.0010 | — | — | — | — | — | — | — | — | — | |
| 8 | GRACE | 0.8746
±
0.0020 | — | — | — | — | — | — | — | — | — | |
| 9 | GAT | 0.8693
±
0.0030 | — | — | — | — | — | — | — | — | — | |
| 10 | GCN | 0.8651
±
0.0050 | — | — | — | — | — | — | — | — | — | |
| 11 | SimGRACE | 0.8604
±
0.0020 | — | — | — | — | — | — | — | — | — | |
| 12 | DGI | 0.8395
±
0.0050 | — | — | — | — | — | — | — | — | — | |
| 13 | GAP-EDP | 0.8380
±
0.0026 | — | — | — | — | — | — | — | — | — | |
| 14 | COLES | 0.7964
±
0.0000 | — | — | — | — | — | — | — | — | — | |
| 15 | DeGTA | 0.7849
±
0.0029 | — | — | — | — | — | — | — | — | — | |
| 16 | NTFormer | 0.7803
±
0.0044 | — | — | — | — | — | — | — | — | — | |
| 17 | NAGphormer | 0.7798
±
0.0016 | — | — | — | — | — | — | — | — | — | |
| 18 | NAGphormer | 0.7743
±
0.0024 | — | — | — | — | — | — | — | — | — | |
| 19 | GAP-NDP | 0.7740
±
0.0007 | — | — | — | — | — | — | — | — | — | |
| 20 | ANS-GT | 0.7632
±
0.0038 | — | — | — | — | — | — | — | — | — | |
| 21 | GRAND | 0.7583
±
0.0021 | — | — | — | — | — | — | — | — | — | |
| 22 | GRAND | 0.7549
±
0.0011 | — | — | — | — | — | — | — | — | — | |
| 23 | GraphSAINT | 0.7521
±
0.0015 | — | — | — | — | — | — | — | — | — | |
| 24 | SGFormer | 0.7422
±
0.0036 | — | — | — | — | — | — | — | — | — | |
| 25 | MLP-DP | 0.7360
±
0.0005 | — | — | — | — | — | — | — | — | — | |
| 26 | NodeFormer | 0.7156
±
0.0042 | — | — | — | — | — | — | — | — | — | |
| 27 | SAGE-EDP | 0.6830
±
0.0099 | — | — | — | — | — | — | — | — | — | |
| 28 | PPRGo | 0.6612
±
0.0059 | — | — | — | — | — | — | — | — | — | |
| 29 | Polynormer | 0.5481
±
0.0049 | — | — | — | — | — | — | — | — | — | |
| 30 | FSGNN | 0.5274
±
0.0083 | — | — | — | — | — | — | — | — | — | |
| 31 | SCNode | 0.5257
±
0.0042 | — | — | — | — | — | — | — | — | — | |
| 32 | CWN | 0.5190
±
0.0015 | — | — | — | — | — | — | — | — | — | |
| 33 | AST | 0.5050
±
0.0027 | — | — | — | — | — | — | — | — | — | |
| 34 | CCCN | 0.5026
±
0.0017 | — | — | — | — | — | — | — | — | — | |
| 35 | GAT | 0.5017
±
0.0059 | — | — | — | — | — | — | — | — | — | |
| 36 | GCN | 0.4956
±
0.0055 | — | — | — | — | — | — | — | — | — | |
| 37 | GIN | 0.4916
±
0.0102 | — | — | — | — | — | — | — | — | — | |
| 38 | UniGNN2 | 0.4906
±
0.0008 | — | — | — | — | — | — | — | — | — | |
| 39 | EDGNN | 0.4818
±
0.0009 | — | — | — | — | — | — | — | — | — | |
| 40 | SGFormer | 0.4818
±
0.0071 | — | — | — | — | — | — | — | — | — | |
| 41 | GCN | 0.4794
±
0.0062 | — | — | — | — | — | — | — | — | — | |
| 42 | RevGAT+T | 0.4722 | — | — | — | — | — | — | — | — | — | |
| 43 | LEMP+T | 0.4675 | — | — | — | — | — | — | — | — | — | |
| 44 | SAGE | 0.4648 | — | — | — | — | — | — | — | — | — | |
| 45 | SAGE+T | 0.4639 | — | — | — | — | — | — | — | — | — | |
| 46 | RevGAT | 0.4590 | — | — | — | — | — | — | — | — | — | |
| 47 | LEMP | 0.4590 | — | — | — | — | — | — | — | — | — | |
| 48 | G2T-LimiX (FT) | 0.4567
±
0.0035 | — | — | — | — | — | — | — | — | — | |
| 49 | GCN | 0.4558 | — | — | — | — | — | — | — | — | — | |
| 50 | GPR | 0.4554 | — | — | — | — | — | — | — | — | — |
SOTA performance
set a new SOTA at the time
matched the SOTA at the time