Mathematical Reasoning on AIME 24 (Acc., Rounds, #Samples)
73.3AccuracySC@32
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SC@32Model=Qwen3-4B-Instruct-Thinking2026.06 | 73.3 | 1 | 32 | |
| ASCModel=Qwen3-4B-Instruct-Thinking2026.06 | 73.3 | 13.8 | 13.8 | |
| RL-guidedModel=Qwen3-4B-Instruct-Thinking2026.06 | 73 | 2.5 | 10 | |
| ESCModel=Qwen3-4B-Instruct-Thinking2026.06 | 72.7 | 3.6 | 18 | |
| SC@32Model=Qwen3-1.7B-Thinking2026.06 | 68.3 | 1 | 32 | |
| ASCModel=Qwen3-1.7B-Thinking2026.06 | 68.2 | 17.7 | 17.7 | |
| RL-guidedModel=Qwen3-1.7B-Thinking2026.06 | 67.6 | 2.9 | 11 | |
| ESCModel=Qwen3-1.7B-Thinking2026.06 | 67.3 | 4.7 | 23.6 | |
| SC@32Model=GPT-4.1-nano2026.06 | 37.1 | 1 | 32 | |
| ASCModel=GPT-4.1-nano2026.06 | 37.1 | 20.6 | 20.6 | |
| ESCModel=GPT-4.1-nano2026.06 | 36.9 | 5.2 | 25.8 | |
| RL-guidedModel=GPT-4.1-nano2026.06 | 36.9 | 6.5 | 16.4 | |
| SC@32Model=Qwen3-0.6B-Thinking2026.06 | 22.6 | 1 | 32 | |
| ASCModel=Qwen3-0.6B-Thinking2026.06 | 22.6 | 27.1 | 27.1 | |
| RL-guidedModel=Qwen3-0.6B-Thinking2026.06 | 22.6 | 5.5 | 21.9 | |
| ESCModel=Qwen3-0.6B-Thinking2026.06 | 22.3 | 6.1 | 30.7 |