For example, we expect the O(LambdaH^2) term in 1loop-beta function for LambdaH to be 32*LambdaH^2 just by replacing the 2HDM couplings in the 2HDM beta functions (the way they are defined in the SARAH model files) by appropriate GTHDM couplings. However, the RGE output for GTHDM is giving 88*Lambda...

Are the RGE for the quartic couplings all accurate in case of the GTHDM? There appears to be a discrepancy when compared to known models (2HDM, Type-II seesaw) in the appropriate limit.