Quality Evaluation on Software reasoning tasks
35.2Functional CorrectnessForce Strong
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Force StrongRouting Strategy=Force Strong2026.01 | 35.2 | 26 | 17.8 | 8.5 | |
| CASTERRouting Strategy=CASTER2026.01 | 34.5 | 25.2 | 16.1 | 9.2 | |
| Force WeakRouting Strategy=Force Weak2026.01 | 31.8 | 25.2 | 16.5 | 8.5 |