Non-linear specification (variable is not a valid distributional or non-linear parameter)

Interfaces brms
1

I am trying to use non-linear syntax to estimate several parameters in a model with brms where I am treating the outcome as a Beta proportion and trying to infer several other parameters.

The general specification is:

\begin{align*} S_{r,t} & \sim Beta(\mu, \kappa) \\ \mu & = I \\ I_{p,d} & = 1 + B_{p,d} \\ B_{p,d} & = a * R^b + c \\ \end{align*}

S_{r,t}, C, and R are observed, so I want to estimate I, B, a, b, and c

I first tried

bform <- bf(
  Srt ~ I,
  I ~ 1 + (C * B )+ (1|r) + (1|t),
  B ~ 1 + (a * R^b + c),
  a ~ 1,
  b ~ 1,
  c ~ 1,
  nl = TRUE
  )

But, that gives me a power error:
Error in terms.formula(as.formula(x)) : invalid power in formula

I assumed this was do to how R is interpreting the ^

So I eliminated b and assumed a 2

bform <- bf(
  seroprev ~ I,
  I ~ 1 + (C * B + (1|r) + (1|t),
  B ~ 1 + (a * R^2 + c),
  a ~ 1,
  c ~ 1,
  nl = TRUE
  )

This gives a different error:
Error: The parameter 'B' is not a valid distributional or non-linear parameter. Did you forget to set 'nl = TRUE'?

I have set nl = TRUE.

Is it possible to estimate multiple parameters this way? Or should I be “injecting” these parameters into the transformed parameters block somehow?