In the annot files generated by mri_glmfit-sim in 5.2.0 (e.g., cache.th20.pos.sig.ocn.annot), the names of the clusters take the form:

cluster0

cluster-001

cluster-002

…

but the last one is of the form

cluster6

which is to say, no dash and no zero-padding.

I assume they are meant to have a consistent format.

Not a big deal, but consistency would be helpful for scripts.

Cheers!

Joe