Math on Minerva
89.16AccuracyCOSE
Evaluation results
| Method | Links | |
|---|---|---|
| COSEBackbone=Llama-3.2-3B-Instruct2026.05 | 89.16 | |
| COSEBackbone=Qwen3-4B2026.05 | 88.38 | |
| MAEBackbone=Qwen3-4B2026.05 | 81.8 | |
| R-ZeroBackbone=Qwen3-4B2026.05 | 81.2 | |
| AZRBackbone=Qwen3-4B2026.05 | 81.1 | |
| BaseBackbone=Qwen3-4B2026.05 | 78.7 | |
| MAEBackbone=Llama-3.2-3B-Instruct2026.05 | 70.2 | |
| R-ZeroBackbone=Llama-3.2-3B-Instruct2026.05 | 69.1 | |
| AZRBackbone=Llama-3.2-3B-Instruct2026.05 | 68.4 | |
| BaseBackbone=Llama-3.2-3B-Instruct2026.05 | 64.1 | |
| Lowest CentroidModel=Qwen3-14B, Number of Samples=642026.04 | 58.1 | |
| Lowest CentroidModel=QWQ-32B, Number of Samples=642026.04 | 57 | |
| Pass@1Model=Qwen3-14B, Number of Samples=642026.04 | 56.3 | |
| Lowest CentroidModel=Olmo-3.1-Think, Number of Samples=642026.04 | 55.9 | |
| Bottom WindowModel=Qwen3-14B, Number of Samples=642026.04 | 55.9 | |
| Bottom WindowModel=QWQ-32B, Number of Samples=642026.04 | 55.9 | |
| Greedy DecodingModel=Qwen3-14B, Number of Samples=642026.04 | 55.5 | |
| Tail ConfidenceModel=Qwen3-14B, Number of Samples=642026.04 | 55.5 | |
| Self-CertaintyModel=Qwen3-14B, Number of Samples=642026.04 | 55.2 | |
| Pass@1Model=Olmo-3.1-Think, Number of Samples=642026.04 | 55 | |
| Greedy DecodingModel=Olmo-3.1-Think, Number of Samples=642026.04 | 54.8 | |
| Bottom WindowModel=Olmo-3.1-Think, Number of Samples=642026.04 | 54.8 | |
| Tail ConfidenceModel=QWQ-32B, Number of Samples=642026.04 | 54.8 | |
| Pass@1Model=QWQ-32B, Number of Samples=642026.04 | 54.5 | |
| Self-CertaintyModel=Olmo-3.1-Think, Number of Samples=642026.04 | 54.4 | |
| Tail ConfidenceModel=Olmo-3.1-Think, Number of Samples=642026.04 | 54.4 | |
| Self-CertaintyModel=QWQ-32B, Number of Samples=642026.04 | 54 | |
| Greedy DecodingModel=QWQ-32B, Number of Samples=642026.04 | 53.3 | |
| Lowest CentroidModel=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 50.4 | |
| Greedy DecodingModel=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 48.2 | |
| Self-CertaintyModel=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 48.2 | |
| Pass@1Model=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 46.7 | |
| Bottom WindowModel=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 46.3 | |
| Tail ConfidenceModel=Ministral-3-14B-Instruct, Number of Samples=642026.04 | 46 | |
| Lowest CentroidModel=GPT-OSS-120B, Number of Samples=642026.04 | 40.1 | |
| Pass@1Model=GPT-OSS-120B, Number of Samples=642026.04 | 38.6 | |
| Tail ConfidenceModel=GPT-OSS-120B, Number of Samples=642026.04 | 38.2 | |
| Bottom WindowModel=GPT-OSS-120B, Number of Samples=642026.04 | 38.2 | |
| Greedy DecodingModel=GPT-OSS-120B, Number of Samples=642026.04 | 37.9 | |
| Self-CertaintyModel=GPT-OSS-120B, Number of Samples=642026.04 | 36 |