Tool Calling on ToolBench Generalization (I1-Tool)
57.7SoPRToolLlama*
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ToolLlama*Setting=Retrieval2026.01 | 57.7 | 48.73 | |
| GPT-3.5*Setting=Retrieval2026.01 | 57.59 | 46.2 | |
| ToolGen*2026.01 | 56.54 | 40.51 | |
| ToolWeaver2026.01 | 54.85 | 36.08 | |
| GPT-4o-miniSetting=Retrieval2026.01 | 53.16 | 49.37 | |
| ToolGen2026.01 | 45.36 | 32.91 | |
| ToolLlama-2Setting=Retrieval2026.01 | 28.48 | 26.58 |