supp_tokenizer_bench / results /js_vs_random_matrix.csv
nancyH's picture
Add Supp. Fig. 2b JS-distance analysis: code, region/background BEDs, FASTAs, results, baseline tokenizers
89dec6b verified
Raw History Blame Contribute Delete
385 Bytes
,promoter,enhancer,exon
DNAbert2,0.2700467663332958,0.1530054987812049,0.16576869998910815
Grover,0.24645834442494668,0.12922598364042234,0.14397131314322673
cCRE_region_BPE,0.2676769058033193,0.1526687523936217,0.17054039887560096
motif_region_BPE,0.2574520985639099,0.1443754257311266,0.16263857681147068
DNAMotifTokenizer,0.20469228698318975,0.10486649424832019,0.12271872972266289