| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| AI Agent, Human, and Crawler Identification Dataset 30-fold cross-validation (test) | MARK (Extra Trees) | Accuracy100 | 5 | 1mo ago | |
| Phantom TTT 500 agents (eval) | Trajectory | Top-1 Accuracy97.5 | 4 | 22d ago | |
| Liar’s Dice (500 eval agents) | Trajectory | Top-1 Accuracy89.5 | 4 | 22d ago | |
| Leduc Poker 500 eval agents | Trajectory | Top-1 Accuracy57.8 | 4 | 22d ago | |
| Kuhn Poker 500 eval agents | Trajectory | Top-1 Accuracy47.9 | 4 | 22d ago | |
| Leduc Poker (eval) | Trajectory Encoder | Top-1 Accuracy57.8 | 4 | 22d ago | |
| Kuhn Poker (eval) | Trajectory Encoder | Top-1 Accuracy47.9 | 4 | 22d ago |