Tool Retrieval on ToolBench In-domain (I3)
91.74NDCG@1IterFeedback
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| IterFeedbackImplementation source=Original2026.01 | 91.74 | 87.94 | 90.2 | |
| ToolWeaverImplementation source=This Paper2026.01 | 88 | 85.8 | 90.12 | |
| ToolGenImplementation source=Original2026.01 | 87 | 85.59 | 90.16 | |
| ToolGenImplementation source=Re-implemented2026.01 | 87 | 85.35 | 90.08 | |
| ToolWeaverImplementation source=This Paper2026.01 | 86 | 86.13 | 90.39 | |
| ToolGenImplementation source=Re-implemented2026.01 | 81 | 80.83 | 85.83 | |
| ToolGenImplementation source=Original2026.01 | 79 | 79.8 | 84.79 | |
| ToolRetrieverImplementation source=Original2026.01 | 70 | 60.44 | 64.7 | |
| ToolRetrieverImplementation source=Re-implemented2026.01 | 70 | 70.01 | 77.21 | |
| Re-InvokeImplementation source=Original2026.01 | 59.65 | — | 59.55 | |
| EmbSimImplementation source=Re-implemented2026.01 | 54 | 46.56 | 52.91 | |
| EmbSimImplementation source=Original2026.01 | 53 | 46.4 | 52.73 | |
| ToolRetrieverImplementation source=Original2026.01 | 52 | 39.89 | 42.92 | |
| BM25Implementation source=Original2026.01 | 32 | 25.88 | 29.78 | |
| BM25Implementation source=Re-implemented2026.01 | 32 | 25.88 | 29.78 | |
| ToolRetrieverImplementation source=Re-implemented2026.01 | 28 | 39.28 | 44.54 | |
| EmbSimImplementation source=Original2026.01 | 18 | 17.77 | 20.7 | |
| EmbSimImplementation source=Re-implemented2026.01 | 18 | 17.77 | 20.94 | |
| BM25Implementation source=Original2026.01 | 10 | 10.08 | 12.33 | |
| BM25Implementation source=Re-implemented2026.01 | 10 | 10.08 | 12.33 |