Quality Estimation on EN-DE Gender-explicit
99.7AccuracyFairQE (ours, w/ COMETKiwi 22)
Evaluation Results
| Method | Links | |
|---|---|---|
| FairQE (ours, w/ COMETKiwi 22)Backbone=COMETKiwi 222026.04 | 99.7 | |
| COMETKiwi 22Backbone=COMETKiwi, Model Scale=222026.04 | 99.2 | |
| COMETKiwi 23 XLBackbone=COMETKiwi, Model Scale=23 XL2026.04 | 98.7 | |
| MetricX 24 XLBackbone=MetricX, Model Scale=24 XL2026.04 | 98.5 | |
| FairQE (ours, w/ MetricX 24 L)Backbone=MetricX 24 L2026.04 | 98.2 | |
| MetricX 24 LBackbone=MetricX, Model Scale=24 L2026.04 | 97.6 | |
| GEMBA-MQMBackbone=GEMBA-MQM2026.04 | 94 |