Reasoning failure prediction and recovery on SAFIM L3
81Accuracythought-tree-based classifier
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| thought-tree-based classifierTarget Reasoning Model=R1, Evaluation Protocol=Generalization from L1 training2026.04 | 81 | 17 | |
| thought-tree-based classifierTarget Reasoning Model=QwQ, Evaluation Protocol=Generalization from L1 training2026.04 | 71 | 4 |