Multi-Turn Tool-Integrated Reasoning (TIR) on AMC23
79.45Peak avg@32 ScoreOTB
Evaluation Results
| Method | Links | |
|---|---|---|
| OTBBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 79.45 | |
| SimpleTIRBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 71.25 | |
| OPOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 64.84 | |
| OGBBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 62.5 | |
| RLOOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 62.42 | |
| GRPOBackbone=Qwen2.5-7B, Reasoning Paradigm=TIR2026.02 | 59.45 |