Task Planning on TaskBench Daily Life
97.36Node-F1Direct
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| DirectLLM=GPT-4-turbo2024.05 | 97.36 | 84.58 | 3.37 | |
| Greedy SearchLLM=GPT-4-turbo2024.05 | 97.36 | 85.78 | 7.37 | |
| BeamSearchLLM=GPT-4-turbo2024.05 | 97.35 | 85.78 | 8.99 | |
| SGCLLM=GPT-4-turbo2024.05 | 97.35 | 85.76 | 3.16 | |
| AdaptiveSearchLLM=GPT-4-turbo2024.05 | 97.3 | 85.8 | 9.04 | |
| SGCLLM=GPT-3.5-turbo2024.05 | 96.38 | 86.19 | 3.16 | |
| DirectLLM=GPT-3.5-turbo2024.05 | 96.09 | 83.65 | 3.36 | |
| BeamSearchLLM=GPT-3.5-turbo2024.05 | 95.66 | 82.72 | 22.05 | |
| AdaptiveSearchLLM=GPT-3.5-turbo2024.05 | 93.79 | 81.41 | 8.53 | |
| Greedy SearchLLM=GPT-3.5-turbo2024.05 | 93.77 | 81.26 | 7.36 | |
| SGCLLM=CodeLlama 13B2024.05 | 92.96 | 79.57 | 3.64 | |
| BeamSearchLLM=CodeLlama 13B2024.05 | 91.74 | 76.6 | 19.19 | |
| Greedy SearchLLM=CodeLlama 13B2024.05 | 91.5 | 76.56 | 5.54 | |
| AdaptiveSearchLLM=CodeLlama 13B2024.05 | 91.34 | 76.09 | 7.18 | |
| DirectLLM=CodeLlama 13B2024.05 | 91.2 | 76.07 | 3.88 | |
| SGCLLM=Mistral 7B2024.05 | 87.13 | 66.49 | 3.54 | |
| AdaptiveSearchLLM=Mistral 7B2024.05 | 86.39 | 63.65 | 10.92 | |
| BeamSearchLLM=Mistral 7B2024.05 | 85.87 | 61.53 | 39.16 | |
| DirectLLM=Mistral 7B2024.05 | 84.26 | 53.63 | 3.77 | |
| SGCLLM=Vicuna 13B2024.05 | 83.33 | 63.77 | 3.82 | |
| GreedySearchLLM=Mistral 7B2024.05 | 75.18 | 49.47 | 8.27 | |
| DirectLLM=Vicuna 13B2024.05 | 73.7 | 45.8 | 3.82 | |
| AdaptiveSearchLLM=Vicuna 13B2024.05 | 62.71 | 31.15 | 13.92 | |
| BeamSearchLLM=Vicuna 13B2024.05 | 54.55 | 23.6 | 24.86 | |
| Greedy SearchLLM=Vicuna 13B2024.05 | 42.98 | 13.33 | 7.18 |