Quality Evaluation on Data reasoning tasks
37.9CorrectnessForce Weak
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Force WeakRouting Strategy=Force Weak2026.01 | 37.9 | 27.6 | 17.8 | 9.9 | |
| Force StrongRouting Strategy=Force Strong2026.01 | 37.4 | 27.1 | 16.8 | 10 | |
| CASTERRouting Strategy=CASTER2026.01 | 37.4 | 27.4 | 17.4 | 9.9 |