| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Wason's rule discovery game | Think-in-Opposites (TiO) | Task Success Rate78 | 33 | 3mo ago | |
| Blicket Task | Qwen3-32B | Task Success Rate86 | 18 | 3mo ago | |
| Blicket baseline (test) | Qwen3-32B | Task Success86 | 8 | 3mo ago | |
| monk2 | NLN | Recovered Rate53.3 | 3 | 27d ago | |
| tic-tac-toe | NLN | Rule Recovery Rate100 | 3 | 27d ago |