Deviation from SS and absolute deviation from 0

Hi everyone,
I have an basic question,
I am writing a model in non-linear form and let Dynare do the log-linearizing job. I am not sure how I should define a variable in “deviation form steady state value” value and another variable in “absolute deviation from 0”.
Please help me out here.

You mean log-deviations vs. absolute deviations? The answer is: put all variables to be log-linearized (i.e. in percentage deviations from steady state) in exp() to perform a variable substitution, see [Model in exp-logs). Variables in absoluted deviations from steady state are not put into exp()