salamandra-7B: benchmark results
Provider: BSC. Released 2024-09-30. Access: Open.
Unified ELO 1311 ± 20, rank #1435 of 1632 rated models, from 342 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LA Leaderboard - xStoryCloze Catalan | 72.67 | Accuracy (%) | 98.5 |
| LA Leaderboard - xStoryCloze Galician | 72.8 | Accuracy (%) | 98.5 |
| LA Leaderboard - BHTCv2 Basque | 46.28 | F1 (%) | 97.1 |
| LA Leaderboard - XCOPA Basque | 61.2 | Accuracy (%) | 97.1 |
| LA Leaderboard - xStoryCloze Basque | 64.86 | Accuracy (%) | 97.1 |
| LA Leaderboard - CatCoLA | 16.94 | MCC (x100) | 95.6 |
| LA Leaderboard - xStoryCloze Spanish | 73.46 | Accuracy (%) | 95.6 |
| LA Leaderboard - PIQA Catalan | 69.64 | Accuracy (%) | 94.1 |
| LA Leaderboard - WiC Basque | 50.93 | Accuracy (%) | 92.6 |
| LA Leaderboard - caBREU Catalan | 18.35 | BLEU | 92.6 |
| LA Leaderboard - COPA Catalan | 80 | Accuracy (%) | 91.2 |
| LA Leaderboard - EusTrivia Basque | 47.99 | Accuracy (%) | 91.2 |
Interactive version: theaggregate.ai/model?slug=salamandra-7b · How It Works · Data refreshed daily, snapshot 2026-10-08.