Manipulation Detection on DeFi Discourse-based Manipulation Dataset
0.9PrecisionHide-and-Shill
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| Hide-and-ShillFramework components=LLM + MARL2025.07 | 0.9 | 0.91 | 0.9 | 0.93 | |
| Deepseek-Detection2025.07 | 0.72 | 0.75 | 0.73 | 0.78 | |
| GCN-Baseline2025.07 | 0.71 | 0.69 | 0.7 | 0.76 | |
| LSTM-Sentiment2025.07 | 0.68 | 0.71 | 0.69 | 0.74 | |
| Rule-Based2025.07 | 0.55 | 0.62 | 0.58 | 0.63 |