nb-roberta-base-scandi-1e4: benchmark results
NB-AiLab nb-roberta-base-scandi-1e4 is a xlm-roberta masked-language checkpoint published by the National Library of Norway lab. Its exact checkpoint label is retained for evaluation and downstream adaptation. Provider: National Library of Norway. Access: Open.
Aggregate Rating 1652 ± 41, rank #608 of 10916 rated models, from 29 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Faroese NLU - FONE | 89.53 | Micro F1 excluding MISC (%) | 99.5 |
| EuroEval Norwegian NLU - NorNE NB | 90.09 | Micro F1 excluding MISC (%) | 99 |
| EuroEval Norwegian NLU - NorNE NN | 85.32 | Micro F1 excluding MISC (%) | 98.2 |
| EuroEval Norwegian NLU - ScaLA NN | 65.86 | Matthews correlation coefficient (%) | 97 |
| EuroEval Icelandic NLU - MIM-GOLD NER | 79.94 | Micro F1 excluding MISC (%) | 96.3 |
| EuroEval Swedish | 72.85 | Average Score (%) | 95.1 |
| EuroEval Norwegian | 67.53 | Average Score (%) | 94.7 |
| EuroEval Swedish NLU | 72.85 | NLU Average Score (%) | 93.3 |
| EuroEval Swedish NLU - SUC3 | 76.09 | Micro F1 excluding MISC (%) | 92.9 |
| EuroEval Swedish NLU - ScaLA SV | 68.16 | Matthews correlation coefficient (%) | 92.9 |
| EuroEval Norwegian NLU | 67.53 | NLU Average Score (%) | 92.7 |
| EuroEval Danish NLU - Dansk | 66.92 | Micro F1 excluding MISC (%) | 91.9 |
Interactive version: theaggregate.ai/model?slug=nb-roberta-base-scandi-1e4 · How It Works · Data refreshed daily, snapshot 2026-11-08.