Algorithmic Problem Solving on self dataset
67.26AC RateMAS-Algorithm
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| MAS-AlgorithmModel=Qwen3-235B-A22B-Instruct-2507, Source=dashscope2026.05 | 67.26 | 67.89 | 71.49 | 0.2 | 10.12 | |
| Direct AskModel=Qwen3-235B-A22B-Instruct-2507, Source=dashscope2026.05 | 62.87 | 63.1 | 65.98 | 0.24 | 8.09 | |
| MAS-AlgorithmModel=Qwen3-Coder-480B-A35B-Instruct, Source=huggingface2026.05 | 52.82 | 52.3 | 56.91 | 0.24 | 13.39 | |
| Direct AskModel=Qwen3-Coder-480B-A35B-Instruct, Source=huggingface2026.05 | 46.49 | 46.41 | 49.71 | 0.54 | 7.74 | |
| MAS-AlgorithmModel=Qwen3-Coder-30B-A3B-Instruct, Source=huggingface2026.05 | 38.87 | 38.51 | 42.28 | 0.16 | 11.18 | |
| MAS-AlgorithmModel=Qwen3-14B, Source=dashscope2026.05 | 37.98 | 40.14 | 42.51 | 0.18 | 20.01 | |
| Direct AskModel=Qwen3-Coder-30B-A3B-Instruct, Training=LoRA Fituning2026.05 | 33.14 | 33.3 | 36.72 | 0.18 | 5.33 | |
| MAS-AlgorithmModel=Qwen3-8B, Source=dashscope2026.05 | 32.25 | 34.16 | 36.41 | 0.07 | 7.61 | |
| Direct AskModel=Qwen3-Coder-30B-A3B-Instruct, Source=huggingface2026.05 | 32.25 | 33.12 | 36.03 | 0.13 | 4.57 | |
| Direct AskModel=Qwen3-14B, Source=dashscope2026.05 | 28.98 | 29.34 | 32.13 | 0.67 | 5.98 | |
| Direct AskModel=Qwen3-8B, Source=dashscope2026.05 | 26.21 | 26.41 | 29.22 | 0.48 | 5.71 | |
| MAS-AlgorithmModel=Qwen3-Coder-30B-A3B-Instruct, Training=LoRA Fituning2026.05 | 23.84 | 25.89 | 27.87 | 0.23 | 6.51 |