LumiOpen/GPQA-FI

--- dataset_info: features: - name: Question dtype: string - name: Correct Answer dtype: string - name: Incorrect Answer 1 dtype: string - name: Incorrect Answer 2 dtype: string - name: Incorrect Answer 3 dtype: string - name: High-level domain dtype: string - name: Subdomain dtype: string - name: Question (pure MT) dtype: string - name: Correct Answer (pure MT) dtype: string - name: Incorrect Answer 1 (pure MT) dtype: string - name: Incorrect Answer 2 (pure MT) dtype: string - name: Incorrect Answer 3 (pure MT) dtype: string splits: - name: train num_bytes: 257629 num_examples: 198 download_size: 174239 dataset_size: 257629 configs: - config_name: default data_files: - split: train path: data/train-* license: mit language: - fi size_categories: - n<1K --- # Dataset Card for GPQA Diamond Finnish This is a manual, human translation GPQA Diamond from Idavidrein/gpqa. the Translation was made by Lingsoft. ### Data Fields - `Question`: The question text - `Correct Answer`: The correct answer - `Incorrect Answer 1-3`: Three distractor answers All data fields also exist in an automatically translated version with the suffix (pure MT) # Citation If you use this resouce, cite the original version and also acknowledge the translation: ```bibtex @misc{gpqa_diamond_fi, author={LUMIOpen}, title={GPQA Diamond Finnish}, year={2025}, url={https://huggingface.co/datasets/LumiOpen/GPQA-FI} } ```