Mathematical Reasoning on AIME 2024 (pass@8)
52.5Pass@8VI-CuRL
Evaluation Results
| Method | Links | |
|---|---|---|
| VI-CuRLBackbone=Qwen2.5-Math-7B, Reward Setting=Oracle (w. Verifier)2026.02 | 52.5 | |
| No CurriculumBackbone=Qwen2.5-Math-7B, Reward Setting=Oracle (w. Verifier)2026.02 | 50 | |
| AdaRFTBackbone=Qwen2.5-Math-7B, Reward Setting=Oracle (w. Verifier)2026.02 | 50 | |
| VCRLBackbone=Qwen2.5-Math-7B, Reward Setting=Oracle (w. Verifier)2026.02 | 48.3 | |
| VI-CuRLBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 41 | |
| BaseBackbone=Qwen2.5-Math-7B, Reward Setting=No RL2026.02 | 36.2 | |
| AdaRFTBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 35.4 | |
| VI-CuRLBackbone=Qwen2.5-Math-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 35.2 | |
| No CurriculumBackbone=Qwen2.5-Math-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 33.3 | |
| No CurriculumBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 32.9 | |
| BaseBackbone=Qwen2.5-Math-1.5B, Reward Setting=No RL2026.02 | 32.4 | |
| VCRLBackbone=Qwen2.5-Math-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 32.2 | |
| AdaRFTBackbone=Qwen2.5-Math-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 32.1 | |
| VCRLBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Reward Setting=Oracle (w. Verifier)2026.02 | 31.9 | |
| BaseBackbone=DeepSeek-R1-Distill-Qwen-1.5B, Reward Setting=No RL2026.02 | 28.7 | |
| No CurriculumBackbone=Llama3.2-3B-Instruct, Reward Setting=Oracle (w. Verifier)2026.02 | 22.4 | |
| VI-CuRLBackbone=Llama3.2-3B-Instruct, Reward Setting=Oracle (w. Verifier)2026.02 | 21.7 | |
| BaseBackbone=Llama3.2-3B-Instruct, Reward Setting=No RL2026.02 | 18.8 | |
| AdaRFTBackbone=Llama3.2-3B-Instruct, Reward Setting=Oracle (w. Verifier)2026.02 | 18.8 | |
| VCRLBackbone=Llama3.2-3B-Instruct, Reward Setting=Oracle (w. Verifier)2026.02 | 16.1 |