List your company

Cross-language Vocabulary Overlap on 6 languages (Same script split)

0.62JSDUnigram
0.60480.70740.810.9126May 26, 2023

Evaluation results

MethodLinks
UnigramTokenizer=Unigram2023.050.62
TokMixTokenizer=TokMix2023.050.65
BPETokenizer=BPE2023.050.68
NoOverlapTokenizer=NoOverlap2023.051