Question Answering on Average (NQ, TabMWP, HotpotQA, TriviaQA, TriState)
38.42EMGreedy
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GreedyModel=Mistral-7B-Inst2026.06 | 38.42 | 49.49 | |
| AdaCADModel=Mistral-7B-Inst2026.06 | 38.13 | 49.11 | |
| ARRModel=Mistral-7B-Inst2026.06 | 38.02 | 49.19 | |
| COIECDModel=Mistral-7B-Inst2026.06 | 34.82 | 46.44 | |
| ARRModel=Llama3-8B-Inst2026.06 | 33.33 | 44.83 | |
| GreedyModel=Llama3-8B-Inst2026.06 | 30.4 | 41.49 | |
| AdaCADModel=Llama3-8B-Inst2026.06 | 27.9 | 38.92 | |
| COIECDModel=Llama3-8B-Inst2026.06 | 25.47 | 36.8 | |
| CADModel=Mistral-7B-Inst2026.06 | 25.19 | 37.13 | |
| GreedyModel=Llama2-13B-Inst2026.06 | 24.52 | 36.68 | |
| AdaCADModel=Llama2-13B-Inst2026.06 | 24.49 | 36.09 | |
| CoCoAModel=Llama3-8B-Inst2026.06 | 23.43 | 34.53 | |
| COIECDModel=Llama2-13B-Inst2026.06 | 22.83 | 34.61 | |
| ARRModel=Llama2-13B-Inst2026.06 | 22.17 | 34.94 | |
| CoCoAModel=Mistral-7B-Inst2026.06 | 19.78 | 33.64 | |
| CADModel=Llama3-8B-Inst2026.06 | 14.57 | 26.04 | |
| CADModel=Llama2-13B-Inst2026.06 | 11.04 | 24.18 | |
| CoCoAModel=Llama2-13B-Inst2026.06 | 4.34 | 21.67 | |
| CoCoAModel=Qwen2.5-7B-Inst2026.06 | 2.86 | 13.99 | |
| CADModel=Qwen2.5-7B-Inst2026.06 | 2.56 | 12.97 | |
| AdaCADModel=Qwen2.5-7B-Inst2026.06 | 2.52 | 13.9 | |
| COIECDModel=Qwen2.5-7B-Inst2026.06 | 2.48 | 13.57 | |
| GreedyModel=Qwen2.5-7B-Inst2026.06 | 1.44 | 13.15 | |
| ARRModel=Qwen2.5-7B-Inst2026.06 | 1.27 | 12.83 |