List your company

Semantic Textual Similarity on KorSTS (test)

77.43Spearman CorrelationKOMBO_Jamo
59.448464.116768.78573.4533Apr 27, 2026

Evaluation results

MethodLinks
KOMBO_JamoTokenization=Jamo, Vocab Size=1702026.0477.43
BERTTokenization=MorSubword, Vocab Size=32K2026.0477.29
BERTTokenization=Subword, Vocab Size=32K2026.0476.41
BERTTokenization=Character, Vocab Size=2K2026.0476.34
KOMBO_StrokeTokenization=Stroke, Vocab Size=1302026.0475.55
KOMBO_CjiTokenization=Cji, Vocab Size=1362026.0475.28
BERTTokenization=Morpheme, Vocab Size=32K2026.0474.35
BERTTokenization=Jamo, Vocab Size=1702026.0472.96
KOMBO_BTSTokenization=BTS, Vocab Size=1122026.0472.9
BERTTokenization=Stroke, Vocab Size=1302026.0467.62
BERTTokenization=Word, Vocab Size=64K2026.0465.5
BERTTokenization=Cji, Vocab Size=1362026.0464.38
BERTTokenization=BTS, Vocab Size=1122026.0460.14