Shuffle Dyck on Shuffle Dyck (test)
100AccuracyFinetuning + KL
Evaluation results
| Method | Links | |
|---|---|---|
| Finetuning + KL2025.05 | 100 | |
| Tracr Injection2025.05 | 99.9 | |
| No editing2025.05 | 0 |
| Method | Links | |
|---|---|---|
| Finetuning + KL2025.05 | 100 | |
| Tracr Injection2025.05 | 99.9 | |
| No editing2025.05 | 0 |