Korean
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Constituency Parsing | Korean | LF Score88.8 | 7 | |
| Automatic Speech Recognition | Korean (ko) low-resource | CER7.2 | 7 | |
| Meaning Transcription | Korean (test) | CER0.54 | 4 | |
| Surface Transcription | Korean (test) | CER6.76 | 4 | |
| Multilingual Watermarking Detection | Korean ko unsupported language subset | AUC0.961 | 4 | |
| Token Fertility | Korean | Fertility Score2.3 | 4 | |
| Subject-level dysarthria severity classification | Korean (KR) (test) | Balanced Accuracy87.3 | 3 | |
| Tokenization | Korean Math | Avg Tokens per Sample1,297.56 | 3 | |
| Tokenization | Korean Reasoning | Avg Tokens per Sample5,520.06 | 3 | |
| Tokenization | Korean General | Avg Tokens/Sample905.82 | 3 |
Showing 10 of 10 rows