Confidence Estimation on Global-MMLU Spanish es (test)
74AUROCCross-lingual Probe
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| Cross-lingual ProbeBackbone=Qwen 3 8B, Evaluation Protocol=Zero-shot2026.05 | 74 | 52 | 0.23 | 0.21 | |
| Verbalized Conf.Backbone=Qwen 3 8B, Evaluation Protocol=Zero-shot2026.05 | 73 | 55 | 0.52 | 0.54 | |
| Seq. LikelihoodBackbone=Qwen 3 8B, Evaluation Protocol=Zero-shot2026.05 | 70 | 51 | 0.53 | 0.59 | |
| Mass-Mean ProbeBackbone=Qwen 3 8B, Evaluation Protocol=Zero-shot2026.05 | 59 | 50 | 0.41 | 0.12 | |
| P(True)Backbone=Qwen 3 8B, Evaluation Protocol=Zero-shot2026.05 | 41 | 23 | 0.26 | 0.26 |