LA Leaderboard - EpecKorrefBin Basque: leaderboard
Binary coreference decisions on Basque text from the EPEC corpus (EpecKorrefBin). 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 | Gemma 2 9B (IT) | 76.32 | #795 |
| 2 | Gemma 2 9B | 72.4 | #924 |
| 3 | Qwen 2.5 7B | 63.88 | #895 |
| 4 | Gemma 3 4B (IT) | 63.54 | #997 |
| 5 | Qwen3-8B-Base | 62.86 | #626 |
| 6 | Llama 3.1 8B Instruct | 62.18 | #1017 |
| 7 | Llama 3.1 8B | 59.45 | #1130 |
| 8 | Qwen 3 8B | 58.26 | #678 |
| 9 | Gemma 2 2B (IT) | 57.07 | #1328 |
| 10 | salamandra-7B | 55.03 | #1446 |
| 11 | gemma-3-4B-pt Base | 54.68 | #1110 |
| 12 | salamandra-2B | 54.17 | #1520 |
| 13 | EuroLLM-1.7B-Instruct | 54 | #1554 |
| 14 | Mistral-7B-v0.3 | 53.66 | #1321 |
| 15 | Mistral 7B Instruct (v0.3) | 53.49 | #1255 |
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-epeckorrefbin-basque · How It Works · Data refreshed daily, snapshot 2026-11-05.