| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| GDPVal 44 tasks (held-out) | Oracle Rubric (Baseline) | Mean Return0.81 | 8 | 3mo ago | |
| Warehouse | Turns Taken12 | 5 | 1d ago | ||
| Sokoban | Success Rate56.2 | 5 | 1d ago | ||
| Search-R1 | Success Rate75.8 | 5 | 1d ago | ||
| SWE-bench | Success Rate71.9 | 5 | 1d ago | ||
| Scientific Discovery Scenario | CASTER | Average Score95.3 | 3 | 3mo ago |