LA Leaderboard - WNLI Catalan: leaderboard
Catalan translation of WNLI, Winograd-schema pronoun resolution posed as entailment. From the LA Leaderboard, which evaluates models on Spanish, Catalan, Basque and Galician tasks with the lm-evaluation-harness; accuracy (%), higher is better. Each model result file is a separate run.
Metric: Accuracy (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 69 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen3-8B-Base | 77.46 | #626 |
| 2 | Qwen 2.5 7B Instruct | 71.83 | #863 |
| 3 | Gemma 2 9B (IT) | 67.61 | #795 |
| 4 | EuroLLM-1.7B-Instruct | 67.61 | #1554 |
| 5 | Qwen 2.5 3B Instruct | 66.2 | #1159 |
| 6 | Llama 3.1 8B Instruct | 64.79 | #1017 |
| 7 | OLMo-2-1124-7B-Instruct | 64.79 | #1343 |
| 8 | Llama 3.2 3B Instruct | 63.38 | #1327 |
| 9 | Qwen 2.5 3B | 63.38 | #1177 |
| 10 | Yi-1.5-9B | 61.97 | #1215 |
| 11 | SmolLM3-3B | 61.97 | #1239 |
| 12 | occiglot-7B-eu5-instruct | 61.97 | |
| 13 | OLMo-2-1124-7B | 61.97 | #1381 |
| 14 | Qwen 2.5 7B | 59.15 | #895 |
| 15 | Gemma 2 9B | 59.15 | #924 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=la-leaderboard-wnli-catalan · How It Works · Data refreshed daily, snapshot 2026-11-05.