rewire.it
Paper-reported evidence

The impact of tokenizer selection in genomic language models

2025 · Peer-reviewed · journal full text in PMC

Paper-reported results. The values below come from this source, not an independent rewire.it run.

Open primary paper →DOI: 10.1093/bioinformatics/btaf456

Final Bioinformatics journal article Table 2, Caduceus (char) Regulatory MCC 0.778 checked directly; same study also has a bioRxiv manuscript.

1 verified numerical rows · source checked 2026-09-15

Paper-reportedIndependent paper evaluation

Caduceus (character tokens) (3.9M parameter variant)

regulatory sequence classification

Reported score
0.778
Metric
MCC
Dataset / split
genomic benchmark categories · paper benchmark summary

task-category MCC across benchmark datasets

Table 2, Regulatory row, Caduceus (char) MCC columnVerify at source →