TSG Execution on 80 real-world incidents
94.38Success RateStepFly
Evaluation Results
| Method | Links | |
|---|---|---|
| StepFlyModel=GPT-4.12025.10 | 94.38 | |
| StepFlyModel=GPT-4o2025.10 | 92.5 | |
| StepFly-wo-QPPModel=GPT-4.12025.10 | 91.88 | |
| StepFlyModel=Grok-32025.10 | 88.75 | |
| StepFly-wo-QPPModel=Grok-32025.10 | 85.62 | |
| StepFlyModel=GPT-4.1-mini2025.10 | 84.38 | |
| StepFly-wo-QPPModel=GPT-4o2025.10 | 81.88 | |
| TaskWeaverModel=Grok-32025.10 | 80.62 | |
| TaskWeaverModel=GPT-4.12025.10 | 80 | |
| StepFly-wo-QPPModel=GPT-4.1-mini2025.10 | 73.75 | |
| ReactModel=Grok-32025.10 | 72.5 | |
| TaskWeaverModel=GPT-4o2025.10 | 72.5 | |
| ReactModel=GPT-4.12025.10 | 71.25 | |
| TaskWeaverModel=GPT-4.1-mini2025.10 | 68.12 | |
| ReactModel=GPT-4o2025.10 | 66.88 | |
| ReactModel=GPT-4.1-mini2025.10 | 61.25 |