It’s not impossible that having more focused priors helps, but usually that’s the most helpful when you have divergences because the log posterior density is too flat and there’s not enough information from the data + prior.
1 Like