Mathematical Reasoning on Math Reasoning Suite Arithmetic Mean
78.1Average Score (@16)THINC-4B
Evaluation Results
| Method | Links | |
|---|---|---|
| THINC-4BReasoning Type=THINC Student2026.05 | 78.1 | |
| Qwen3-235B-A22B-ThinkingReasoning Type=NL-Only2026.05 | 75.2 | |
| ASTER-4BReasoning Type=Tool-Integrated2026.05 | 73.8 | |
| Qwen3-4B-Thinking-2507Reasoning Type=NL-Only2026.05 | 65 | |
| gpt-oss-20B (medium)Reasoning Type=NL-Only2026.05 | 64.7 | |
| Qwen3.5-27B*Reasoning Type=THINC Teacher, Python Interpreter=prompted2026.05 | 64.7 | |
| OpenReasoning-Nemotron-7BReasoning Type=NL-Only2026.05 | 62.7 | |
| Qwen3-4B-Thinking-2507*Reasoning Type=Tool-Integrated, Python Interpreter=prompted2026.05 | 62.6 | |
| rStar2-Agent-14BReasoning Type=Tool-Integrated2026.05 | 61.5 | |
| DemyAgent-4BReasoning Type=Tool-Integrated2026.05 | 57.3 | |
| ReTool-32BReasoning Type=Tool-Integrated2026.05 | 50.7 | |
| THINC-1.7BReasoning Type=THINC Student2026.05 | 42.8 | |
| Qwen3-1.7BReasoning Type=NL-Only2026.05 | 32.2 | |
| Qwen3-1.7B*Reasoning Type=Tool-Integrated, Python Interpreter=prompted2026.05 | 29.8 | |
| CoRT-1.5BReasoning Type=Tool-Integrated2026.05 | 25.7 |