Mathematical Reasoning on AIME24 search
85.8AccuracyAutoTTS (β=1.0)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AutoTTS (β=1.0)Base Model=Qwen3-8B, Type=Discovered2026.05 | 85.8 | 467,400 | |
| AutoTTS (β=0.5)Base Model=Qwen3-8B, Type=Discovered2026.05 | 84.3 | 255,300 | |
| AutoTTS (β=1.0)Base Model=Qwen3-4B, Type=Discovered2026.05 | 83.5 | 424,900 | |
| AutoTTS (β=0.5)Base Model=Qwen3-4B, Type=Discovered2026.05 | 82 | 236,700 | |
| Parallel-ProbeBase Model=Qwen3-8B, Type=Handcrafted2026.05 | 81.5 | 730,800 | |
| SC @ 64Base Model=Qwen3-8B, Type=Handcrafted2026.05 | 80.4 | 910,800 | |
| ASCBase Model=Qwen3-8B, Type=Handcrafted2026.05 | 80.4 | 226,000 | |
| ESCBase Model=Qwen3-8B, Type=Handcrafted2026.05 | 80.4 | 459,400 | |
| ASCBase Model=Qwen3-4B, Type=Handcrafted2026.05 | 80.1 | 175,700 | |
| SC @ 64Base Model=Qwen3-4B, Type=Handcrafted2026.05 | 80 | 886,800 | |
| ESCBase Model=Qwen3-4B, Type=Handcrafted2026.05 | 80 | 528,900 | |
| Parallel-ProbeBase Model=Qwen3-4B, Type=Handcrafted2026.05 | 79.7 | 688,900 | |
| SC @ 64Base Model=Qwen3-1.7B, Type=Handcrafted2026.05 | 72.5 | 1,025,800 | |
| ESCBase Model=Qwen3-1.7B, Type=Handcrafted2026.05 | 72.5 | 909,200 | |
| ASCBase Model=Qwen3-1.7B, Type=Handcrafted2026.05 | 72.3 | 482,600 | |
| AutoTTS (β=1.0)Base Model=Qwen3-1.7B, Type=Discovered2026.05 | 70.4 | 499,100 | |
| AutoTTS (β=0.5)Base Model=Qwen3-1.7B, Type=Discovered2026.05 | 68.5 | 276,300 | |
| Parallel-ProbeBase Model=Qwen3-1.7B, Type=Handcrafted2026.05 | 68.1 | 748,500 | |
| Parallel-ProbeBase Model=Qwen3-0.6B, Type=Handcrafted2026.05 | 21.8 | 773,800 | |
| SC @ 64Base Model=Qwen3-0.6B, Type=Handcrafted2026.05 | 21.4 | 1,008,600 | |
| ASCBase Model=Qwen3-0.6B, Type=Handcrafted2026.05 | 21.4 | 805,500 | |
| ESCBase Model=Qwen3-0.6B, Type=Handcrafted2026.05 | 21.4 | 986,700 | |
| AutoTTS (β=1.0)Base Model=Qwen3-0.6B, Type=Discovered2026.05 | 20.9 | 542,200 | |
| AutoTTS (β=0.5)Base Model=Qwen3-0.6B, Type=Discovered2026.05 | 19.2 | 283,600 |