Multilingual LLM-as-a-Judge on MMLU-ProX
60.27AccuracySEMJ
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SEMJ2026.06 | 60.27 | 0.5738 | |
| Monolingual Ref2026.06 | 58.49 | 0.4951 | |
| Multilingual Vote2026.06 | 57.14 | — | |
| Base Judge2026.06 | 56.18 | 0.5534 | |
| Monolingual Vote2026.06 | 56.11 | 0.5548 |