Multi-hop reasoning on MuSiQue Open Book (test)
75.05Accuracy / F1DeCoRe (static)
Evaluation Results
| Method | Links | |
|---|---|---|
| DeCoRe (static)Model=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 75.05 | |
| DoLaModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.64 | |
| GreedyModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.43 | |
| ITIModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.27 | |
| ADModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.27 | |
| DCO (Ours)Model=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.27 | |
| CADModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 74.02 | |
| CADModel=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 73.02 | |
| DeCoRe (static)Model=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 72.49 | |
| CDModel=Llama-3-70B, Chain-of-Thought=With CoT2026.06 | 71.7 | |
| DoLaModel=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 69.92 | |
| ADModel=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 69.92 | |
| GreedyModel=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 69.84 | |
| DeCoRe (static)Model=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 69.76 | |
| DCO (Ours)Model=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 69.01 | |
| DoLaModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 68.68 | |
| CADModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 68.64 | |
| GreedyModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 68.56 | |
| ITIModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 68.14 | |
| ADModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 68.14 | |
| DCO (Ours)Model=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 67.89 | |
| CDModel=Llama-3-70B, Chain-of-Thought=Without CoT2026.06 | 66.61 | |
| DeCoRe (static)Model=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 61.23 | |
| DoLaModel=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 59.08 | |
| GreedyModel=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 58.83 | |
| ADModel=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 58.63 | |
| DCO (Ours)Model=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 58.26 | |
| CADModel=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 57.88 | |
| ITIModel=Llama-3-8B, Chain-of-Thought=Without CoT2026.06 | 45.84 | |
| ITIModel=Llama-3-8B, Chain-of-Thought=With CoT2026.06 | 38.31 |