Question Answering on FreshQA (out-of-domain)
67.2PrecisionKP
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| KPBackbone=Gemma-2-9B2026.02 | 67.2 | 61.2 | 64.1 | |
| VeriFY-NMBackbone=Gemma-2-9B2026.02 | 59.5 | 82.2 | 69 | |
| VeriFY-MMBackbone=Gemma-2-9B2026.02 | 58.2 | 87.7 | 69.9 | |
| VeriFY-SMBackbone=Gemma-2-9B2026.02 | 56.4 | 84.1 | 67.5 | |
| SCBackbone=Gemma-2-9B2026.02 | 54.8 | 90.9 | 68.4 | |
| KPBackbone=Gemma-2-2B2026.02 | 54.1 | 69.4 | 60.8 | |
| BaseBackbone=Gemma-2-9B2026.02 | 50 | — | — | |
| VeriFY-MMBackbone=Gemma-2-2B2026.02 | 49.3 | 81.8 | 61.5 | |
| SCBackbone=Gemma-2-2B2026.02 | 46.6 | 82.7 | 59.6 | |
| VeriFY-NMBackbone=Gemma-2-2B2026.02 | 46.1 | 68.6 | 55.1 | |
| VeriFY-SMBackbone=Gemma-2-2B2026.02 | 45.2 | 78.6 | 57.4 | |
| BaseBackbone=Gemma-2-2B2026.02 | 37.3 | — | — |