Mathematical Reasoning on Minerva (Pass@1, # Tokens)
31.76Pass@1 Success RateLACONIC
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LACONICBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=LACONIC, Length Budget (B)=20002026.02 | 31.76 | 1,869 | |
| GRPOBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=GRPO2026.02 | 29.64 | 2,630 | |
| ThinkPrune-Iter2kBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=ThinkPrune-Iter2k2026.02 | 29.24 | 1,941 | |
| DeepSeek-1.5BBase Model=DeepSeek-1.5B2026.02 | 29.13 | 6,783 | |
| L1-ExactBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=L1-Exact2026.02 | 28.74 | 3,755 | |
| LACONICBase Model=DeepSeek-1.5B, Fine-tuning Algorithm=LACONIC, Length Budget (B)=15002026.02 | 28.36 | 1,638 | |
| L1-MaxBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=L1-Max2026.02 | 28.35 | 1,701 | |
| DeepScaleR-1.5BBase Model=DeepScaleR-1.5B2026.02 | 27.62 | 4,134 | |
| Efficient-ReasoningBase Model=DeepSeek-1.5B, Fine-tuning Algorithm=Efficient-Reasoning2026.02 | 27.57 | 2,018 | |
| ThinkPrune-Iter2kBase Model=DeepSeek-1.5B, Fine-tuning Algorithm=ThinkPrune-Iter2k2026.02 | 27.35 | 1,837 | |
| GRPOBase Model=DeepSeek-1.5B, Fine-tuning Algorithm=GRPO2026.02 | 27.2 | 2,655 | |
| Efficient-ReasoningBase Model=DeepScaleR-1.5B, Fine-tuning Algorithm=Efficient-Reasoning2026.02 | 27.11 | 1,852 |