Mathematical Reasoning on AIME 2024 (ave@32)
77.6Accuracy (ave@32)Qwen3-4B + Direct-OPD
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-4B + Direct-OPDTransfer source=JustRL policy-shift2026.07 | 77.6 | |
| Qwen3-4BTransfer source=JustRL policy-shift2026.07 | 72.5 | |
| Teacher RLTransfer source=QuestA policy-shift2026.07 | 72.5 | |
| R1-Distill-7B + Direct-OPDTransfer source=JustRL policy-shift2026.07 | 63.1 | |
| Qwen3-1.7B + Direct-OPDTransfer source=JustRL policy-shift2026.07 | 62.4 | |
| Teacher refTransfer source=QuestA policy-shift2026.07 | 61.77 | |
| R1-Distill-7B + Direct-OPDTransfer source=QuestA policy-shift2026.07 | 61.2 | |
| Qwen3-1.7B + Direct-OPDTransfer source=QuestA policy-shift2026.07 | 59 | |
| R1-Distill-7BTransfer source=JustRL policy-shift2026.07 | 56.7 | |
| R1-Distill-7BTransfer source=QuestA policy-shift2026.07 | 56.3 | |
| Teacher RLTransfer source=JustRL policy-shift2026.07 | 51.3 | |
| Qwen3-1.7BTransfer source=QuestA policy-shift2026.07 | 49.1 | |
| Qwen3-1.7BTransfer source=JustRL policy-shift2026.07 | 48.3 | |
| Teacher refTransfer source=JustRL policy-shift2026.07 | 28.5 |