Tool-use gating on When2Call
99.23TC AccuracyInit
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| InitModel=Gemma3-4B2026.05 | 99.23 | 5.56 | 57.11 | |
| AMCSModel=Gemma3-4B2026.05 | 99.18 | 5.55 | 57 | |
| InitModel=Gemma3-1B2026.05 | 98.61 | 2.26 | 55.2 | |
| PromoteModel=Gemma3-4B2026.05 | 98.53 | 7.64 | 57.58 | |
| PromoteModel=Gemma3-1B2026.05 | 98.38 | 2.97 | 55.39 | |
| PromoteModel=Qwen3.5-9B2026.05 | 96.88 | 34.27 | 68.67 | |
| InitModel=Qwen3.5-9B2026.05 | 96.45 | 37.57 | 69.88 | |
| InitModel=Qwen3.5-4B2026.05 | 95.98 | 37.76 | 69.75 | |
| PromptModel=Gemma3-1B2026.05 | 94.67 | 15.91 | 59.19 | |
| SuppressModel=Qwen3.5-9B2026.05 | 94.61 | 44.43 | 72 | |
| SuppressModel=Gemma3-4B2026.05 | 94.34 | 7.65 | 55.28 | |
| SuppressModel=Gemma3-1B2026.05 | 93.84 | 4.35 | 53.52 | |
| PromoteModel=Qwen3.5-4B2026.05 | 93.82 | 47.36 | 72.89 | |
| SuppressModel=Qwen3.5-4B2026.05 | 93.48 | 44.65 | 71.48 | |
| InitModel=Ministral3-8B2026.05 | 91.97 | 41.71 | 69.33 | |
| AMCSModel=Qwen3.5-4B2026.05 | 91.82 | 51.07 | 73.47 | |
| AMCSModel=Qwen3.5-9B2026.05 | 91.72 | 54.58 | 74.98 | |
| InitModel=Ministral3-3B2026.05 | 91.66 | 34.65 | 65.97 | |
| PromoteModel=Ministral3-3B2026.05 | 90.07 | 41.26 | 68.08 | |
| PromoteModel=Ministral3-8B2026.05 | 89.92 | 46.91 | 70.54 | |
| AMCSModel=Ministral3-8B2026.05 | 89.9 | 45.76 | 70.1 | |
| AMCSModel=Gemma3-1B2026.05 | 89.77 | 9.4 | 53.56 | |
| SuppressModel=Ministral3-3B2026.05 | 86.75 | 45.75 | 68.27 | |
| AMCSModel=Ministral3-3B2026.05 | 86.66 | 45.91 | 68.3 | |
| SuppressModel=Ministral3-8B2026.05 | 85.34 | 52.75 | 70.66 | |
| PromptModel=Qwen3.5-4B2026.05 | 68.11 | 77.31 | 72.25 | |
| PromptModel=Qwen3.5-9B2026.05 | 67.72 | 79.38 | 72.97 | |
| PromptModel=Ministral3-3B2026.05 | 61.62 | 71.56 | 66.1 | |
| PromptModel=Gemma3-4B2026.05 | 58.22 | 59.23 | 58.68 | |
| PromptModel=Ministral3-8B2026.05 | 35.91 | 87.1 | 58.97 |