Tool Usage on Meta-Tool Multiple
91.1AccuracyTHINKBRAKE
Evaluation results
| Method | Links | ||
|---|---|---|---|
| THINKBRAKEBase Model=Qwen3-4B, Stopping Strategy=THINKBRAKE2025.10 | 91.1 | -3 | |
| NoWaitBase Model=Phi-4-Reasoning, Stopping Strategy=NoWait2025.10 | 90.9 | -15.4 | |
| Qwen3-4BBase Model=Qwen3-4B, Stopping Strategy=Base2025.10 | 90.5 | — | |
| NoWaitBase Model=Qwen3-4B, Stopping Strategy=NoWait2025.10 | 90.5 | -8.4 | |
| Phi-4-ReasoningBase Model=Phi-4-Reasoning, Stopping Strategy=Base2025.10 | 90.5 | — | |
| THINKBRAKEBase Model=Phi-4-Reasoning, Stopping Strategy=THINKBRAKE2025.10 | 90.3 | -5.4 | |
| THINKBRAKE-pBase Model=Qwen3-4B-Thinking, Stopping Strategy=THINKBRAKE-p2025.10 | 86.7 | -12.7 | |
| ThinkLessBase Model=Qwen3-4B-Thinking, Stopping Strategy=ThinkLess2025.10 | 86.5 | -100 | |
| THINKBRAKEBase Model=Qwen3-4B-Thinking, Stopping Strategy=THINKBRAKE2025.10 | 86.3 | -9 | |
| THINKBRAKEBase Model=Qwen3-14B, Stopping Strategy=THINKBRAKE2025.10 | 85.7 | -4.5 | |
| Qwen3-4B-ThinkingBase Model=Qwen3-4B-Thinking, Stopping Strategy=Base2025.10 | 85.5 | — | |
| THINKBRAKEBase Model=Qwen3-32B, Stopping Strategy=THINKBRAKE2025.10 | 85.5 | -1.1 | |
| Qwen3-14BBase Model=Qwen3-14B, Stopping Strategy=Base2025.10 | 84.9 | — | |
| NoWaitBase Model=Qwen3-4B-Thinking, Stopping Strategy=NoWait2025.10 | 84.7 | -11.2 | |
| NoWaitBase Model=Qwen3-14B, Stopping Strategy=NoWait2025.10 | 84.5 | -5.8 | |
| Qwen3-32BBase Model=Qwen3-32B, Stopping Strategy=Base2025.10 | 84.3 | — | |
| NoWaitBase Model=Qwen3-32B, Stopping Strategy=NoWait2025.10 | 83.9 | 1.3 | |
| DeepSeek-R1-7BBase Model=DeepSeek-R1-7B, Stopping Strategy=Base2025.10 | 80.3 | — | |
| THINKBRAKEBase Model=DeepSeek-R1-7B, Stopping Strategy=THINKBRAKE2025.10 | 76.3 | -26.3 | |
| NoWaitBase Model=DeepSeek-R1-7B, Stopping Strategy=NoWait2025.10 | 73.2 | -19.3 |