Open-ended Generation on NQ entity-swapped (test)
73.73Exact MatchHICD
Evaluation Results
| Method | Links | |
|---|---|---|
| HICDBackbone=Mistral-7B-v0.32025.03 | 73.73 | |
| CADBackbone=Mistral-7B-v0.32025.03 | 73.07 | |
| HICDBackbone=LLaMA3-8B-Instruct2025.03 | 72.64 | |
| CADBackbone=LLaMA3-8B-Instruct2025.03 | 72.51 | |
| HICDBackbone=LLaMA-7B2025.03 | 69.61 | |
| CADBackbone=LLaMA-7B2025.03 | 68.24 | |
| DoLABackbone=Mistral-7B-v0.32025.03 | 62.24 | |
| Mistral-7B-v0.3Backbone=Mistral-7B-v0.32025.03 | 61.41 | |
| DoLABackbone=LLaMA3-8B-Instruct2025.03 | 58.86 | |
| LLaMA3-8B-InstructBackbone=LLaMA3-8B-Instruct2025.03 | 58.74 | |
| LLaMA-7BBackbone=LLaMA-7B2025.03 | 56.25 | |
| DoLABackbone=LLaMA-7B2025.03 | 56.14 |