Session-level validation on URS (test)
0.6952MAETask-guarded evaluator
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Task-guarded evaluatorCategory=Ours2026.05 | 0.6952 | 0.276 | 0.2883 | 0.2424 | 52.07 | |
| Calibrated memory promptCategory=Memory baseline2026.05 | 0.726 | 0.2358 | 0.2495 | 0.2073 | 45.92 | |
| Calibrated promptCategory=No memory2026.05 | 0.7928 | 0.1621 | 0.1527 | 0.1411 | 34.15 | |
| Structured memory promptCategory=Memory baseline2026.05 | 0.8305 | 0.2312 | 0.2354 | 0.1868 | 53.86 | |
| Direct promptCategory=No memory2026.05 | 1.0497 | 0.1588 | 0.1295 | 0.1396 | 51.52 |