It depends on my mood - If my silly switch is activated I entertain and run with ideas I know to be incorrect. Otherwise I let the idea pop up, show it around the Si storehouse and if it falls short I discount it. However I do give myself a moment to see if I can extract any nuggets from the aforementioned failed ideas which may have potential, which weren't completely contradictory to known data and ones which I think could branch out into new possibilities and ideas which have more potential. However this isn't always the case as I am aware the Si isn't perfect itself so if I later find the Si to have had faulty stock I often re-consider previously dismissed possibilities.
In any case the best policy for my Ne is to approach it with an open mind and a healthy degree of scepticism and uncertainty...but not so open and uncertain that it becomes a hole in the floor which I fall into.