Mathematical Reasoning on AIME 24 (Pass@8, Pass@16, Pass@32)
63.33Pass@8SENT
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SENTBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 63.33 | 70 | 76.67 | |
| w/ EnBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 60 | 60 | 70 | |
| w/ MaskBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 56.67 | 63.33 | 73.33 | |
| w/ AdvBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 53.33 | 56.67 | 53.33 | |
| w/ ClipBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 46.67 | 46.67 | 70 | |
| w/ High EnBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 46.67 | 60 | 53.33 | |
| w/ CovBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 43.33 | 53.33 | 53.33 | |
| Base modelBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 40 | 43.33 | 43.33 | |
| GRPOBackbone=DeepSeek-R1-Distill-Qwen-1.5B2025.12 | 40 | 50 | 56.67 |