Should the repressor really have a negative correlation with its motif?

[jase11]

Chromvar is a good package to evaluate the motif activity in scATAC-seq data, but I noticed some studies using chromvar have said that there should be a negative correlation between the expression of a repressor and its motif activity, which confused me.

In my opinion, whether the transcription factor is an activator or a repressor, it could regulate the downstream target only when it binds to the cis-regulatory element containing its motif. Therefore, even if the transcription factor is a repressor, there should be a positive correlation between its expression and the accessibility of the cis-regulatory element containing its motif.

Still, I'm not sure whether the negative correlation between a transcription factor and its motif can be a basis for regarding the transcription factor as a repressor, as I did see some cases of the same in my own data. Does the z-score of motif activity calculated by chromvar really reflect the accessibility of the corresponding genomic region, or is it a measure of something else?

I would be appreciated if you could clear my confusion.