Natural Language Inference on SNLI (Macro-F1)
72.59Macro-F1MATCHA
Evaluation Results
| Method | Links | |
|---|---|---|
| MATCHAThreshold=fixed midpoint threshold2026.05 | 72.59 | |
| R1-F1Threshold=fixed midpoint threshold2026.05 | 57.69 | |
| RL-F1Threshold=fixed midpoint threshold2026.05 | 54.67 | |
| METEORThreshold=fixed midpoint threshold2026.05 | 46.68 | |
| BLEURTThreshold=fixed midpoint threshold2026.05 | 44.72 | |
| R2-F1Threshold=fixed midpoint threshold2026.05 | 42.35 | |
| SimCSEThreshold=fixed midpoint threshold2026.05 | 39.2 | |
| EmbSimThreshold=fixed midpoint threshold2026.05 | 37.96 | |
| BERTScoreThreshold=fixed midpoint threshold2026.05 | 33.33 |