Mathematical Reasoning on AIME 24 (MV/WMV Acc, Tokens)
89.2Accuracy (MV)DeepConf
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DeepConfBackbone=DeepSeek-8B, branching factor n=322026.06 | 89.2 | 90 | 16.3 | |
| ASCBackbone=DeepSeek-8B, branching factor n=322026.06 | 86 | — | 10.1 | |
| Flex-Fork-thinkBackbone=DeepSeek-8B, branching factor n=32, seed path length l0=2,0482026.06 | 86 | 86.6 | 9.8 | |
| DeepConfBackbone=Qwen3-8B, branching factor n=322026.06 | 81.8 | 81.2 | 11.1 | |
| Flex-Fork-thinkBackbone=Qwen3-8B, branching factor n=32, seed path length l0=2,0482026.06 | 81.8 | 81.8 | 7 | |
| ASCBackbone=Qwen3-8B, branching factor n=322026.06 | 80.1 | — | 10.4 | |
| DeepConfBackbone=Phi-4-RP-14B, branching factor n=322026.06 | 80 | 80 | 8.9 | |
| ASCBackbone=Phi-4-RP-14B, branching factor n=322026.06 | 77.8 | — | 6.4 | |
| Flex-Fork-thinkBackbone=Phi-4-RP-14B, branching factor n=32, seed path length l0=2,0482026.06 | 77 | 76.6 | 6.7 |