Clarification Seeking on ALFWorld Clarification (test)
68.9Success RateUncertainty Decomposition
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Uncertainty DecompositionModel=GLM-4.72026.06 | 68.9 | 73.7 | 60.9 | 66.7 | 68.9 | |
| ReAct+UEModel=GLM-4.72026.06 | 64.7 | 98.5 | 4.5 | 8.7 | 58.8 | |
| UAMModel=DeepSeek-v3.2-exp2026.06 | 56 | 84.8 | 70 | 76.7 | 79.8 | |
| UAMModel=GLM-4.72026.06 | 54.7 | 81.8 | 25 | 38.3 | 61.3 | |
| ReAct+UEModel=DeepSeek-v3.2-exp2026.06 | 54 | 67.5 | 90 | 77.1 | 74.6 | |
| UAMModel=GPT-5.12026.06 | 47.1 | 73.7 | 42.4 | 53.8 | 64.7 | |
| Uncertainty DecompositionModel=GPT-5.12026.06 | 45.7 | 93.8 | 48.4 | 63.8 | 75.7 | |
| ReAct+UEModel=GPT-5.12026.06 | 36.2 | 64.3 | 25.7 | 36.7 | 55.1 | |
| Uncertainty DecompositionModel=GPT-OSS-120B2026.06 | 28.4 | 71.4 | 72.9 | 72.2 | 71.6 | |
| ReAct+UEModel=GPT-OSS-120B2026.06 | 27.1 | 55.6 | 20.4 | 29.9 | 51 | |
| UAMModel=GPT-OSS-120B2026.06 | 26.1 | 50 | 15.9 | 24.1 | 52.2 | |
| ReAct+UEModel=Qwen3.5-35B2026.06 | 23.9 | 76.5 | 30.2 | 43.3 | 52.1 | |
| Uncertainty DecompositionModel=DeepSeek-v3.2-exp2026.06 | 22.9 | 59.1 | 92.9 | 72.2 | 63.9 | |
| UAMModel=Qwen3.5-35B2026.06 | 21.4 | 65.7 | 50 | 56.8 | 58.3 | |
| Uncertainty DecompositionModel=Qwen3.5-35B2026.06 | 20.6 | 56.1 | 74.2 | 63.9 | 61.8 |