Attacking Cascade System Decision Maker on SQuAD 2.0
86.5Performance
LLM w/o cascade
Updated 15d ago
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
2026.05 | 86.5 | 277.24 | 8.49 | 100 | 100 | |
2026.05 | 81 | 110.31 | 4.97 | 51.5 | 24 | |
2026.05 | 70.5 | 242.17 | 8.27 | 61.5 | 50.5 | |
2026.05 | 69 | 141.89 | 7.04 | 52 | 24.5 | |
2026.05 | 68 | 206.4 | 7.64 | 60.5 | 42 | |
2026.05 | 63 | 216.3 | 7.62 | 59.5 | 41.5 | |
2026.05 | 60 | 174.98 | 6.41 | 54 | 38.5 | |
2026.05 | 52.5 | 191.74 | 6.72 | 55 | 32 | |
2026.05 | 43.5 | 158.54 | 5.89 | 54 | 24 | |
2026.05 | 36 | 131.33 | 5.05 | 52 | 15 | |
2026.05 | 28.5 | 90.89 | 4.37 | 54.5 | 8.5 |