Does anyone know if the natural language parsing is showing improvement, or does it still choke on most inputs?
9 | | 73 | | 241 | | 561 | | 1081 | | 1849
| 64 | | 168 | | 320 | | 520 | | 768 |
| | 104 | | 152 | | 200 | | 248 | |
| | | 48 | | 48 | | 48 | | |