Task Goal Completion on AppWorld Easy normal 57 tasks (test)
73.7AccuracyLRE
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| LRE2026.06 | 73.7 | 5.97 | 0 | |
| FIFOstrategy=recency2026.06 | 59.6 | 4.98 | 0 | |
| No compression2026.06 | 57.9 | 6.51 | 0 | |
| ACONmode=in-fork2026.06 | 57.9 | 4.67 | 3.19 |
| Method | Links | |||
|---|---|---|---|---|
| LRE2026.06 | 73.7 | 5.97 | 0 | |
| FIFOstrategy=recency2026.06 | 59.6 | 4.98 | 0 | |
| No compression2026.06 | 57.9 | 6.51 | 0 | |
| ACONmode=in-fork2026.06 | 57.9 | 4.67 | 3.19 |