Open PL LLM - DYK (multiple choice, 5-shot): leaderboard
Did You Know: deciding whether a candidate answer to a Polish Wikipedia question is correct. Open PL LLM Leaderboard (SpeakLeash), 5-shot, multiple choice variant: the options are ranked by log-likelihood; binary F1 (%), higher is better. Runs not available on the Hugging Face Hub are excluded.
Metric: Binary F1 (%). Source: huggingface.co. Saturation forecast: Around February 2027. 326 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Mistral Large 2 (Nov) Instruct (2411) | 74.61 | #480 |
| 2 | Qwen 2.5 72B Instruct | 73.81 | #469 |
| 3 | Mistral Large 2 (Jul) | 73.8 | #490 |
| 4 | Qwen 2.5 32B | 73.03 | #544 |
| 5 | Qwen 2.5 72B | 71.58 | #512 |
| 6 | cogito-v1-preview-qwen-14B | 71.46 | |
| 7 | Mixtral 8x22B Instruct | 70.93 | #738 |
| 8 | QwQ 32B-Preview | 70.89 | #671 |
| 9 | Llama 3.1 405B Instruct FP8 | 70.67 | #400 |
| 10 | Bielik-11B-v2.3-Instruct | 70.26 | #700 |
| 11 | Bielik-11B-v2.2-Instruct | 70.24 | #669 |
| 12 | Qwen 2 72B Instruct | 70.08 | #579 |
| 13 | Bielik-11B-v2.1-Instruct | 70.05 | #679 |
| 14 | Qwen 3 32B | 69.61 | #449 |
| 15 | Llama 3.3 70B Instruct | 69.41 | #541 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=open-pl-llm-dyk-multiple-choice-5-shot · How It Works · Data refreshed daily, snapshot 2026-11-05.