Quality Estimation on EN-* Gender-ambiguous Fem. vs. Masc.
99.47QE Score (ES)FairQE
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| FairQEBackbone=COMETKiwi 222026.04 | 99.47 | 98.57 | 99.17 | 99.38 | 99.93 | 99.65 | |
| FairQELLM Evaluator=gpt-4.1-mini, QE Backbone=COMETKiwi 222026.04 | 99.47 | 98.57 | 99.17 | 99.38 | 99.93 | 99.65 | |
| FairQEBackbone=MetricX 24 L2026.04 | 98.76 | 97.31 | 98.81 | 96.5 | 99.54 | 99.56 | |
| COMETKiwi 222026.04 | 98.32 | 97.83 | 97.91 | 98.51 | 99.37 | 99.09 | |
| MetricX 24 L2026.04 | 98.04 | 97.14 | 97.82 | 96.23 | 99.11 | 99.45 | |
| MetricX 24 XL2026.04 | 98.02 | 97.01 | 98.16 | 99.43 | 99.86 | 99.89 | |
| GEMBA-MQMLLM Evaluator=o32026.04 | 97.38 | 97.23 | 96.6 | 97.67 | 97.72 | 97.62 | |
| GEMBA-MQM2026.04 | 97.37 | 96.58 | 96.95 | 97 | 97.49 | 97.4 | |
| GEMBA-MQMLLM Evaluator=gpt-4.1-mini2026.04 | 97.37 | 96.58 | 96.95 | 97 | 97.49 | 97.4 | |
| COMETKiwi 23 XL2026.04 | 93.98 | 90.28 | 92.61 | 98.41 | 99.06 | 98.4 |