MultiWikiQHalluA - Icelandic: leaderboard
Metric: Answer-level hallucination rate (%): share of the 1,000 Icelandic MultiWikiQA test answers in which the fine-tuned mmBERT-small LettuceDetect token classifier flags at least one token as unsupported by the Wikipedia context, times 100; lower is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 5 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 3 14B | 26 |
| 2 | Gemma 3 12B (IT) | 27 |
| 3 | Qwen 3 0.6B | 60 |
Interactive version: theaggregate.ai/benchmark?slug=multiwikiqhallua-icelandic · How It Works · Data refreshed daily, snapshot 2026-10-07.