| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MIB Arithmetic (test) | Full DAS | Accuracy98.4 | 8 | 2mo ago | |
| Formal-language benchmark lengths 41-500 (test) | TRNN | Accuracy100 | 6 | 2mo ago | |
| Addition (OOD lengths 20-60) | RL-Halt | OOD Performance45 | 4 | 25d ago | |
| add2 binary addition (test) | - | - | 0 | 5mo ago |