Reasoning failure prediction and recovery on SAFIM L1
79Accuracythought-tree-based classifier
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| thought-tree-based classifierTarget Reasoning Model=R12026.04 | 79 | 39 | |
| thought-tree-based classifierTarget Reasoning Model=QwQ2026.04 | 79 | 16 |