Semantic Uncertainty Discrimination on SQuAD
0.78E-SCTS
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TSEvaluation Protocol=SEconf, Model=Qwen2026.04 | 0.78 | 0.704 | 0.772 | 0.78 | |
| SEEvaluation Protocol=SEconf, Model=Qwen, Temperature Scaling=tau=0.52026.04 | 0.666 | 0.655 | 0.669 | 0.665 | |
| PlattEvaluation Protocol=SEconf, Model=Qwen2026.04 | 0.665 | 0.649 | 0.66 | 0.666 | |
| ATSEvaluation Protocol=SEconf, Model=Qwen2026.04 | 0.579 | 0.649 | 0.649 | 0.591 | |
| BaseEvaluation Protocol=SEconf, Model=Qwen, Temperature Scaling=tau=1.02026.04 | 0.569 | 0.605 | 0.598 | 0.571 |