SILMA 9B: benchmark results
Provider: SILMA AI. Access: Open.
Aggregate Rating 1494 ± 35, rank #2462 of 10916 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Arabic - Alghafa - Multiple Choice Facts Truefalse Balanced Task | 97.33 | Exact Match (%) | 97.5 |
| HELM Arabic - Alghafa - Multiple Choice Rating Sentiment No Neutral Task | 88.9 | Exact Match (%) | 97.5 |
| HELM Arabic - Alghafa - Multiple Choice Rating Sentiment Task | 68.3 | Exact Match (%) | 90 |
| HELM Arabic Enterprise - Arabic Finance Bool | 77.19 | EM | 85.3 |
| HELM Arabic Enterprise - Arabic Finance Mcq | 91.3 | EM | 80.9 |
| HELM Arabic - Arabic Exams - Biology | 42.86 | Exact Match (%) | 80 |
| HELM Arabic - Alghafa - Multiple Choice Sentiment Task | 42.2 | Exact Match (%) | 56.2 |
| HELM Arabic - Aratrust - Privacy | 96.49 | Exact Match (%) | 50 |
| HELM Arabic - Alghafa | 75.84 | EM | 46.2 |
| HELM Arabic - Arabic Mmlu - General Knowledge (Middle School) | 77.91 | Exact Match (%) | 40 |
| HELM Arabic - Alghafa - Multiple Choice Grounded Statement Soqal Task | 88 | Exact Match (%) | 37.5 |
| HELM Arabic - Mbzuai Human Translated Arabic Mmlu - College Chemistry | 48 | Exact Match (%) | 33.8 |
Interactive version: theaggregate.ai/model?slug=silma-9b · How It Works · Data refreshed daily, snapshot 2026-11-08.