salamandra-2B-instruct: benchmark results
Provider: BSC. Released 2024-09-30. Access: Open.
Unified ELO 1222 ± 26, rank #1945 of 2096 rated models, from 75 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| LA Leaderboard - XNLI Basque | 42.03 | Accuracy (%) | 88.2 |
| LA Leaderboard - WNLI Basque | 57.75 | Accuracy (%) | 87.5 |
| LA Leaderboard - xStoryCloze Basque | 54.53 | Accuracy (%) | 83.8 |
| LA Leaderboard - BHTCv2 Basque | 26.65 | F1 (%) | 76.5 |
| LA Leaderboard - XNLI Galician | 46.84 | Accuracy (%) | 75 |
| LA Leaderboard - caBREU Catalan | 13.45 | BLEU | 75 |
| LA Leaderboard - PIQA Catalan | 64.31 | Accuracy (%) | 69.1 |
| LA Leaderboard - OpenBookQA Galician | 26.6 | Accuracy (%) | 68.4 |
| LA Leaderboard - TE-ca | 47.85 | Accuracy (%) | 67.6 |
| LA Leaderboard - COPA Catalan | 70.6 | Accuracy (%) | 61.8 |
| LA Leaderboard - RagQuAS | 77.44 | Semantic Answer Similarity (x100) | 61.8 |
| LA Leaderboard - XCOPA Basque | 51.8 | Accuracy (%) | 61.8 |
Interactive version: theaggregate.ai/model?slug=salamandra-2b-instruct · How It Works · Data refreshed daily, snapshot 2026-10-08.