LGAI-EXAONE/KMMLU-Redux
KMMLU-Redux π Paper We introduce KMMLU-Redux, a reconstructed version of the existing KMMLU, comprising 2,587 problems from Korean National Technical Qualification (KNTQ) exams. We identified several critical issues in the KMMLU, including leaked answers, lack of clarity, ill-posed questions, notation errors, and contamination risks. To address these problems, we conducted rigorous manual examination and decontaminated the dataset from the pre-training corpus to preventβ¦ See the full description on the dataset page: https://huggingface.co/datasets/LGAI-EXAONE/KMMLU-Redux.
| Author / organization | LGAI-EXAONE |
|---|---|
| License (as declared) | cc-by-nc-nd-4.0 |
| Downloads | 280 (as reported by Hugging Face) |
| Likes | 26 |
| Created | 11 July 2025 |
| Last updated at source | 15 July 2025 |
Popularity, last 3 days
280 now (+10 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 16 Oct 2026 Description: KMMLU-Redux π Paper We introduce KMMLU-Redux, a reconstructed version of the existing KMMLU, comprising 2,587 problems from Korean National Technical Qualification (KNTQ) exams. We identified several critical issues in the KMMLU, including β KMMLU-Redux π Paper We introduce KMMLU-Redux, a reconstructed version of the existing KMMLU, comprising 2,587 problems from Korean National Technical Qualification (KNTQ) exams. We identified several critical issues in the KMMLU, including
Changes detected by Silytics between refreshes of the source.
More from LGAI-EXAONE
- LGAI-EXAONE/KMMLU-ProDataset
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.