Nonlinear model with grouped data and a single control, model specification problem in BRMS

I think I must be getting close. Here is another similar query:

posted here: Non-linear mixed effects model specification

Mysteriously the coef call has an added suffix “M” or “P”. But this is the first time i’ve seen fixed effects called in a nonlinear setup. I’ve also trawled through set_prior manual but, although a lot of this is going over my head, I’m not seeing anything there that wouldnt make “normal(xbar,sd”, class”b”, coef=”groupingName”, nlpar=”ref nlpar parameter here”) the logical syntax.

However when I use this syntax, it looks like STAN spits back this to brms:

pretty obvious its croaking on the priors syntax.

Thomas