possible issue with nAGQ>1 ?

[Ben Bolker]

I've come across an example where blme seems to give weird results with nAGQ>1 (i.e., non-trivial Gauss-Hermite quadrature rather than Laplace approximation). I haven't thoroughly explored/gotten together a reproducible example yet, but thought I would post a comment here in the meantime (in case this has been reported before, or someone else finds it later before I have a chance to replicate/go into greater detail).