salamandra-2B — benchmark results
Provider: BSC. Released 2024-09-30. Access: Open.
Unified ELO 1258 ± 52, rank #1604 of 1776 rated models, from 19 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LA Leaderboard - XNLI Galician | 45.9 | Accuracy (%) | 58.8 |
| LA Leaderboard - GalCoLA | 51.61 | Accuracy (%) | 56.6 |
| LA Leaderboard - PIQA Catalan | 62.02 | Accuracy (%) | 53.7 |
| EuroEval Finnish NLU - Scandisent FI | 88.87 | Sentiment classification Score (%) | 48.3 |
| EuroEval Portuguese NLU - SST-2 PT | 71.59 | Sentiment classification Score (%) | 37.7 |
| LA Leaderboard - COPA Spanish | 71.2 | Accuracy (%) | 27.2 |
| EuroEval Dutch NLU - DBRD | 69.02 | Sentiment classification Score (%) | 26.7 |
| EuroEval Italian NLU - ScaLA IT | 2.88 | Linguistic acceptability Score (%) | 24.2 |
| EuroEval Italian NLU - Sentipolc16 | 29.64 | Sentiment classification Score (%) | 19.6 |
| LA Leaderboard - Spanish Law Exams | 25.21 | Accuracy (%) | 16.2 |
| LA Leaderboard | 48.27 | Average Score (%) | 14.7 |
| EuroEval Danish NLU - Angry Tweets | 15.68 | Sentiment classification Score (%) | 13.8 |
Interactive version: theaggregate.ai/model?slug=salamandra-2b · How the rankings work · Data refreshed daily, snapshot 2026-07-22.