Tool-use Planning on ToolBench G3-Instruction
0.9368Win RateGPT4 TOPGUN
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT4 TOPGUNReference=GPT4 ReACT2024.02 | 0.9368 | |
| GPT4 TOPGUNReference=T.LLAMA REACT2024.02 | 0.9323 | |
| GPT4 TOPGUNReference=ChatGPT REACT2024.02 | 0.9005 | |
| GPT4 TOPGUNReference=T.LLAMA DFSDT2024.02 | 0.8947 | |
| GPT4 TOPGUNReference=T.LLAMA DFSDT+Ret2024.02 | 0.8782 | |
| GPT4 TOPGUNReference=ChatGPT DFSDT2024.02 | 0.8526 | |
| GPT4 DFSDTReference=ChatGPT REACT2024.02 | 0.84 | |
| GPT4 TOPGUNReference=GPT4 DFSDT2024.02 | 0.7925 | |
| GPT4 ReACTReference=ChatGPT REACT2024.02 | 0.78 | |
| T.LLAMA DFSDT+RetReference=ChatGPT REACT2024.02 | 0.73 | |
| T.LLAMA DFSDTReference=ChatGPT REACT2024.02 | 0.69 | |
| ChatGPT DFSDTReference=ChatGPT REACT2024.02 | 0.69 | |
| T.LLAMA REACTReference=ChatGPT REACT2024.02 | 0.55 |