Retrieval-Augmented Question Answering on NQ (PIA/PoisonedRAG Evaluation)
89Clean Accuracy
VanillaRAG
Updated 2mo ago
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
2026.05 | 89 | 31 | 65 | 84 | 68 | |
2026.05 | 88 | 21 | 72 | 83 | 78 | |
2026.05 | 87 | 30 | 77 | 85 | 76 | |
2026.05 | 83 | 34 | 81 | 59 | 67 | |
2026.05 | 83 | 31 | 81 | 57 | 68 | |
2026.05 | 80 | 29 | 56 | 91 | 18 | |
2026.05 | 79 | 29 | 80 | 19 | 82 | |
2026.05 | 78 | 34 | 75 | 69 | 62 | |
2026.05 | 77 | 26 | 69 | 89 | 34 | |
2026.05 | 77 | 16 | 60 | 93 | 16 | |
2026.05 | 77 | 38 | 66 | 28 | 75 | |
2026.05 | 76 | 20 | 59 | 87 | 37 | |
2026.05 | 76 | 20 | 59 | 91 | 20 | |
2026.05 | 73 | 29 | 58 | 87 | 44 | |
2026.05 | 72 | 18 | 72 | 15 | 79 | |
2026.05 | 68 | 55 | 49 | 77 | 29 | |
2026.05 | 66 | 29 | 55 | 24 | 65 | |
2026.05 | 65 | 20 | 64 | 10 | 72 | |
2026.05 | 64 | 25 | 60 | 30 | 63 | |
2026.05 | 63 | 51 | 57 | 79 | 24 | |
2026.05 | 63 | 54 | 48 | 79 | 32 | |
2026.05 | 62 | 79 | 49 | 62 | 42 | |
2026.05 | 62 | 25 | 71 | 16 | 62 | |
2026.05 | 62 | 19 | 61 | 6 | 71 | |
2026.05 | 61 | 22 | 64 | 8 | 75 | |
2026.05 | 60 | 74 | 48 | 61 | 40 | |
2026.05 | 60 | 70 | 50 | 62 | 43 | |
2026.05 | 55 | 32 | 50 | 70 | 41 | |
2026.05 | 54 | 40 | 44 | 70 | 53 | |
2026.05 | 48 | 41 | 44 | 61 | 43 | |
2026.05 | 38 | 48 | 39 | 63 | 27 | |
2026.05 | 36 | 75 | 52 | 50 | 45 | |
2026.05 | 31 | 64 | 44 | 46 | 33 | |
2026.05 | 30 | 34 | 48 | 55 | 31 | |
2026.05 | 30 | 79 | 57 | 45 | 33 | |
2026.05 | 25 | 39 | 51 | 54 | 33 | |
2026.05 | 25 | 72 | 59 | 51 | 44 | |
2026.05 | 24 | 64 | 39 | 52 | 42 | |
2026.05 | 22 | 68 | 44 | 50 | 39 | |
2026.05 | 21 | 33 | 15 | 40 | 17 | |
2026.05 | 17 | 33 | 14 | 49 | 18 | |
2026.05 | 16 | 34 | 18 | 45 | 14 | |
2026.05 | 12 | 29 | 17 | 40 | 18 | |
2026.05 | 12 | 28 | 11 | 41 | 15 | |
2026.05 | 9 | 18 | 15 | 35 | 14 |