NUTS and ridges

Thanks for this discussion. This entire thread was very illuminating for understanding how posterior geometries can go wrong because of non-computational problems such as ill-specified models and insufficient or problematic experimental design