Correlation analysis with ground truth on Evaluation Dataset 2000 samples
0.754Pearson Correlation Coefficientsts_paraphrase
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| sts_paraphrasecategory=Best single evaluator2026.03 | 0.754 | 0.806 | |
| median over evaluatorscategory=Best consensus baseline2026.03 | 0.749 | 0.801 | |
| Semantic qualitytype=dimension2026.03 | 0.733 | 0.776 | |
| Compositeweights=default weights2026.03 | 0.513 | 0.682 | |
| Structure qualitytype=dimension2026.03 | 0.466 | 0.601 | |
| Agreement / uncertaintytype=dimension2026.03 | -0.384 | 0.008 | |
| Query-output alignmenttype=dimension2026.03 | -0.437 | -0.238 |