I dunno - this seems like a problem ideally situated for adversarial networks to resolve. All you need is a smaller, dumber, validated model that can check and

I dunno - this seems like a problem ideally situated for adversarial networks to resolve. All you need is a smaller, dumber, validated model that can check and validate the output of the bigger, smarter model, and either use that to directly act as a filter between the user and the larger model, or use it to retrain the larger model until it isn’t dumb and wrong.