Assumption logistic regression: linearity of independent variables and log odds?

I checked whether in my logistic regression models the assumption of linearity between the independent variables and the log odds is met. I used R and followed the instructions on this website to do so: http://www.sthda.com/english/articles/36-classification-methods-essentials/148-logistic-regression-assumptions-and-diagnostics-in-r/

I am estimating several models adjusted and unadjusted; and with and without interaction term.The variables of the interaction term are continuous

For most of the independent variables and most of the estimated models this assumption is met, but for some models I get the following (see picture). I noticed that such an association between the independent variable and the log odds occurs only in models WITH interactionterm and WITHOUT adjustment variables. Can someone explain why this pattern occurs and what it means?

Thank you in advance

Franziska