Choice of regularizing prior for the cumulative probit

[epkanol]

I’m still quite new to Stan and Bayesian modeling in general, but I was following the pattern laid out in a blog post by @Solomon : Notes on the Bayesian cumulative probit | A. Solomon Kurz

Unfortunately, brms does not yet support a more proper Dirichlet prior on the cutpoints:

Dirichlet prior for cutpoints in ordinal models

And I have not yet gotten this workaround to work, so I’m resorting to the separated normal distributions:

Dirichlet prior on ordinal regression cutpoints in brms

In a detailed tutorial, @betanalpha makes the case for using a dirichlet prior in ordinal regression models (Ordinal Regression). I confess, I wasn’t able to understand all that much of it besides the fact that a dirichlet prior sounds like a good idea. Is there any way to implement a weakly informative dirichlet prior on the cutpoints in brms? This was also mentioned by @paul.buerkner as a possible feature to develop (Dirichlet prior for cutpoints in ordinal models · Issue #762 · paul-buerkner…

Most of my data sets sample correctly, even in the face of outliers (where I had to resort to reloo to figure out that one problematic estimated Pareto-k-value was in fact not an issue). But this is conditional on me using exp(5) for the sd.
I have given up on exp(1).