snunlp/KoBALT-700
KoBALT: Korean Benchmark For Advanced Linguistic Tasks KoBALT is designed to advance the evaluation of large language models (LLMs) in Korean, a morphologically rich language, by addressing the limitations of conventional benchmarks that often lack linguistic depth and typological grounding. It consists of 700 expert-written multiple-choice questions covering 24 fine-grained linguistic phenomena across five core linguistic domains: Syntax (300) Semantics (215) Pragmatics (81)… See the full description on the dataset page: https://huggingface.co/datasets/snunlp/KoBALT-700.
| Author / organization | snunlp |
|---|---|
| License (as declared) | cc-by-nc-4.0 |
| Downloads | 927 (as reported by Hugging Face) |
| Likes | 24 |
| Created | 15 May 2025 |
| Last updated at source | 9 September 2025 |
Popularity, last 3 days
927 now (+14 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
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 Information may be incomplete or out of date.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.