Mathematical Reasoning on AIME24 (Full Len., Core Len., CR, RM, Top-3 Mass, Retention)
11.4Full Length ScoreMinimal-core extraction
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Minimal-core extractionModel=GPT-52026.05 | 11.4 | 6 | 53 | 47 | 64 | 85 | |
| Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-32B2026.05 | 10.9 | 6.3 | 58 | 42 | 59 | 83 | |
| Minimal-core extractionModel=Qwen3-32B2026.05 | 10.5 | 6.5 | 62 | 38 | 56 | 81 | |
| Minimal-core extractionModel=DeepSeek-R1-Distill-Qwen-7B2026.05 | 10.2 | 6.7 | 66 | 34 | 53 | 77 |