Mathematical Reasoning on MATH 500 (Pass@1, Avg Response Length)
94.2Pass@1 AccuracyExpThink
Evaluation results
| Method | Links | ||
|---|---|---|---|
| ExpThinkBackbone=Qwen3-8B2026.05 | 94.2 | 755.7 | |
| VanillaBackbone=Qwen3-8B2026.05 | 93 | 5,315.2 | |
| ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 92.4 | 1,253 | |
| LaserBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 91.6 | 1,868.2 | |
| AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 91.3 | 2,109.8 | |
| AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 91 | 1,752.2 | |
| JETBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 90.8 | 2,971.1 | |
| VanillaBackbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 89.8 | 3,723.8 | |
| LC-R1Backbone=DeepSeek-R1-Distill-Qwen-7B2026.05 | 89 | 1,529.3 | |
| LaserBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 84.6 | 2,602.7 | |
| AdaptThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 83.8 | 1,851.1 | |
| JETBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 82.8 | 3,179.6 | |
| ExpThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 82.2 | 1,417 | |
| VanillaBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 81 | 4,635.5 | |
| AutoThinkBackbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 79.8 | 2,358.3 | |
| LC-R1Backbone=DeepSeek-R1-Distill-Qwen-1.5B2026.05 | 78.6 | 2,257.1 |