Tool Retrieval on TOOLRET Zero-Shot Web*
34.9nDCG@10bge-gemma
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| bge-gemmaMethod Category=Re-ranking with cross-encoder2026.01 | 34.9 | 19.5 | |
| TOOLQPMethod Category=Fine-tuning, Training Data Size=10k, Backbone Model=1.7B2026.01 | 33 | 23.1 | |
| bge-m3Method Category=Re-ranking with cross-encoder2026.01 | 32.7 | 18.5 | |
| Q2E/PRFMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 31 | 18.8 | |
| Q2E/ZSMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 30.8 | 19.6 | |
| Q2P/SFTMethod Category=Fine-tuning, Training Data Size=10k, Backbone Model=1.5B/1.7B models2026.01 | 30.7 | 19.7 | |
| Q2D/PRFMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 30.5 | 18.9 | |
| Base Retriever (gte-Qwen)Method Category=Base Retriever2026.01 | 29.7 | 17.8 | |
| gte-Qwen/ContrastiveFTMethod Category=Fine-tuning, Training Data Size=10k, Backbone Model=1.5B/1.7B models2026.01 | 29.1 | 18.6 | |
| Re-InvokeMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 28.1 | 18.5 | |
| D2QMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 26.6 | 16.3 | |
| HyDE/PRFMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 26.5 | 18 | |
| Q2D/ZSMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 24.5 | 18.1 | |
| TOOLQP-FORMATMethod Category=Fine-tuning, Training Data Size=10k, Backbone Model=1.7B2026.01 | 22.6 | 17.6 | |
| HyDE/ZSMethod Category=Prompting, Backbone Model=Qwen3-30B-A3B-Instruct2026.01 | 19.7 | 17 |