Open-Ended Instruction Task Execution on Minecraft Open-Ended Instruction Tasks (test)
75Torch Success RateGOAP
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| GOAPPlanner=GPT-4V, Policy=GOAP2025.02 | 75 | 47 | 13 | 16 | 17 | |
| GOAPPlanner=GLM-4V, Policy=GOAP2025.02 | 71 | 39 | 11 | 14 | 13 | |
| STEVE-1Planner=GPT-4V, Policy=STEVE-12025.02 | 66 | 10 | 0 | 0 | 0 | |
| STEVE-1Planner=GLM-4V, Policy=STEVE-12025.02 | 60 | 0 | 0 | 0 | 0 | |
| VPT (text)Planner=GPT-4V, Policy=VPT (text)2025.02 | 11 | 0 | 0 | 0 | 0 | |
| VPT (text)Planner=GLM-4V, Policy=VPT (text)2025.02 | 5 | 0 | 0 | 0 | 0 |