Research map / Graph learning / Graph self-supervised learning
IGT4ETH: An Isotropic Pre-trained Graph Transformer for Ethereum Account Classification
Ao Liu, Yanmei Zhang, Youwei Wang, Qiang Duan
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use IGT4ETH graph transformer instead of ZipZap | pre-trained model · Exchange (fixed-training) | F1 (%): 79.65 -> 87.13 | Improved |
| add edge embeddings and edge channel (ablation) | IGT4ETH · Mining (fixed-training) | F1 (%): 73.97 -> 82.21 | Improved |
| add structural centrality embeddings (ablation) | IGT4ETH · Exchange (fixed-training) | F1 (%): 80.75 -> 87.13 | Improved |
| add structural role embeddings (ablation) | IGT4ETH · Exchange (fixed-training) | F1 (%): 85.05 -> 87.13 | Improved |
| use Focal-InfoNCE instead of InfoNCE (ablation) | IGT4ETH · Phishing (fixed-training) | F1 (%): 55.22 -> 58.76 | Improved |
| add Conceptor Negation post-processing (ablation) | IGT4ETH · ICO-Wallet (fixed-training) | F1 (%): 82.98 -> 86.32 | Improved |
| add pre-training (ablation) | IGT4ETH · Exchange (fine-tuning) | F1 (%): 77.64 -> 87.92 | Improved |
| min-max scaling post-processing (ablation) | IGT4ETH · Mining (fixed-training) | F1 (%): 82.21 -> 0 | Hurt |
| unit length normalization post-processing (ablation) | IGT4ETH · ICO-Wallet (fixed-training) | F1 (%): 82.98 -> 71.43 | Hurt |
| CN vs abtt, z-score, cluster-based post-processing (ablation) | IGT4ETH · Mining (fixed-training) | F1 (%): abtt 83.33 vs CN 82.76 | Mixed |
Closest papers by meaning
- The Temporal Graph of Bitcoin Transactions (NeurIPS 2025 · similarity 0.83)
- Multi-Chain Graphs of Graphs: A New Approach to Analyzing Blockchain Datasets (NeurIPS 2024 · similarity 0.83)
- EX-Graph: A Pioneering Dataset Bridging Ethereum and X (ICLR 2024 · similarity 0.83)
- BlockScan: Detecting Anomalies in Blockchain Transactions (NeurIPS 2025 · similarity 0.83)
- Context-aware Graph Neural Network for Graph-based Fraud Detection with Extremely Limited Labels (AAAI 2025 · similarity 0.82)
- Provably Powerful Graph Neural Networks for Directed Multigraphs (AAAI 2024 · similarity 0.82)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
