Confidence Estimation on Global-MMLU Russian
73AUROCLast Answer
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Last AnswerProbe hidden state=last answer token, Training Language=French, Test Language=Russian, Zero-shot Transfer=true2026.05 | 73 | 48 | 31 | 35 | |
| Last QueryProbe hidden state=last query token, Training Language=French, Test Language=Russian, Zero-shot Transfer=true2026.05 | 67 | 34 | 23 | 21 |