Mathematical Reasoning on AMC23 (EM and Tool Usage)
77.5EM AccuracyParetoPO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ParetoPOBase Model=Qwen2.5-Math-7B, Tool Use=Yes2026.06 | 77.5 | 1.3 | |
| ToRL-GRPOBase Model=Qwen2.5-Math-7B, Tool Use=Yes2026.06 | 75 | 2.6 | |
| OTC-GRPOBase Model=Qwen2.5-Math-7B, Tool Use=Yes2026.06 | 72.5 | 1.6 | |
| ParetoPOBase Model=Qwen2.5-Math-1.5B, Tool Use=Yes2026.06 | 70 | 0.8 | |
| Qwen2.5-Math-7B-Ins-TIRBase Model=Qwen2.5-Math-7B, Tool Use=Yes2026.06 | 70 | 2.3 | |
| ToRL-GRPOBase Model=Qwen2.5-Math-1.5B, Tool Use=Yes2026.06 | 67.5 | 2.5 | |
| MO-GRPOBase Model=Qwen2.5-Math-7B, Tool Use=Yes2026.06 | 67.5 | 1.7 | |
| Qwen2.5-Math-7B-InsBase Model=Qwen2.5-Math-7B, Tool Use=No2026.06 | 65 | — | |
| Qwen2.5-Math-1.5B-InsBase Model=Qwen2.5-Math-1.5B, Tool Use=No2026.06 | 62.5 | — | |
| OTC-GRPOBase Model=Qwen2.5-Math-1.5B, Tool Use=Yes2026.06 | 62.5 | 1.1 | |
| MO-GRPOBase Model=Qwen2.5-Math-1.5B, Tool Use=Yes2026.06 | 62.5 | 2.1 | |
| Qwen2.5-Math-1.5B-Ins-TIRBase Model=Qwen2.5-Math-1.5B, Tool Use=Yes2026.06 | 55 | 2 |