An optimization I’m doing (for work, not the RNN) consistently gets better results when I turn off an inequality constraint that in practice is never active. The obvious quick fix is to just turn off the constraint, since it’s never relevant anyway. But still, what’s going on here?
