Mathematical Reasoning on OlympicB
57.64AccuracyASLEC-CASL
Evaluation Results
| Method | Links | |
|---|---|---|
| ASLEC-CASLBackbone=Qwen-4B-Instruct2026.04 | 57.64 | |
| ASLEC-DROPBackbone=Qwen-4B-Instruct2026.04 | 57.18 | |
| ASLEC-CASLBackbone=Qwen-8B-Base2026.04 | 51.02 | |
| ASLEC-DROPBackbone=Qwen-8B-Base2026.04 | 49.33 | |
| Local LPBackbone=Qwen-4B-Instruct2026.04 | 49.26 | |
| GRACEBackbone=Qwen-4B-Instruct2026.04 | 48.82 | |
| ASLEC-CASLBackbone=Qwen-4B-Base2026.04 | 48.23 | |
| ASLEC-DROPBackbone=Qwen-4B-Base2026.04 | 47.2 | |
| ASLEC-CASLBackbone=Qwen-7B-Instruct2026.04 | 46.32 | |
| Local LPBackbone=Qwen-8B-Base2026.04 | 44.41 | |
| GRACEBackbone=Qwen-8B-Base2026.04 | 44.11 | |
| Local LPBackbone=Qwen-4B-Base2026.04 | 43.08 | |
| GRACEBackbone=Qwen-4B-Base2026.04 | 42.5 | |
| ASLEC-DROPBackbone=Qwen-7B-Instruct2026.04 | 41.61 | |
| Local LPBackbone=Qwen-7B-Instruct2026.04 | 35.58 | |
| GRACEBackbone=Qwen-7B-Instruct2026.04 | 34.85 |