Mathematical Reasoning on AIME 2024 (Pass@32, Mem, Len)
93Pass@32TFGO
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| TFGOBase Model=DeepSeek-V3.2-Exp2025.12 | 93 | — | 696 | |
| VanillaBase Model=Qwen3-Max2025.12 | 93 | — | — | |
| MNLBase Model=Qwen3-Max2025.12 | 93 | 10 | 0 | |
| MNLBase Model=DeepSeek-V3.2-Exp2025.12 | 90 | 9 | 60 | |
| TFGOBase Model=Qwen3-Max2025.12 | 90 | — | 1,452 | |
| VanillaBase Model=DeepSeek-V3.2-Exp2025.12 | 87 | — | — | |
| ACEBase Model=DeepSeek-V3.2-Exp2025.12 | 80 | 163 | 21,318 | |
| MNLBase Model=Qwen3-8B2025.12 | 33 | 51 | 67 | |
| VanillaBase Model=Qwen3-8B2025.12 | 30 | — | — | |
| ACEBase Model=Qwen3-8B2025.12 | 27 | 100 | 7,355 | |
| TFGOBase Model=Qwen3-8B2025.12 | 23 | — | 703 | |
| MementoBase Model=Qwen3-8B2025.12 | 20 | 100 | 3,100 |